File: ChangeLog

package info (click to toggle)
stacks 2.41+dfsg-1
  • links: PTS, VCS
  • area: main
  • in suites: bullseye, sid
  • size: 2,740 kB
  • sloc: cpp: 37,468; sh: 1,953; perl: 1,174; python: 497; makefile: 139
file content (1404 lines) | stat: -rw-r--r-- 87,830 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
Stacks 2.41 - July 8, 2019
--------------------------
    Feature: populations: calculates haplotype-based Dxy (Nei, 1987) and provides for smoothing if a 
             reference genome is available.
    Feature: populations: re-implemented full sequence export for phylip format, including partitioning information.

Stacks 2.4 - May 9, 2019
------------------------
    Feature: populations: re-implemented HZAR export.
    Feature: added reporting code to detect issues with inconsistent versions of libz on a host system
             causing Stacks components to fail to open compressed files.
    Feature: gstacks: improved PCR duplicate reporting to be per-sample.
    Bugfix:  populations: fixed an issues where the basepair position of a small number of loci was reported
                          incorrectly -- they were shifted by a small, fixed offset.

Stacks 2.3 - Jan 11, 2019
--------------------------
    Feature: populations: Backwards-compatibly worked on filtering options; added long names for -r and -p and
             added --min-samples-overall and --filter-haplotype-wise.
    Feature: populations: Implemented --treemix.
    Feature: gstacks: Improved RAD-loci reference sequences around the end of forward (restriction site-bound) reads.
    Feature: gstacks: Improved the way 2-microsatelittes are dealt with.
    Feature: gstacks: Changed the default value for --var-alpha from 0.05 to 0.01 (--gt-alpha is unchanged at 0.05).
    Feature: gstacks: Improved PCR duplicates-related log information (distribution of clone sizes).
    Feature: gstacks: Added an option to save read alignments (--write-alignments).
    Feature: Backwards-compatibly switched to hyphens in command line options (underscores remain accepted
             where they previously were).
    Feature: cstacks/sstacks now report an error when the disk becomes full.

	2.3b - Jan 23, 2019
	----------
	Bugfix: Fixed some limit cases causing an abort at gstacks.cc:1752.
	Bugfix: Fixed some limit cases causing an abort at debruijn.cc:60.
	Bugfix: Fixed some limit cases apparently causing an infinite loop in the de Bruijn code.
	Bugfix: Restored compilation with the oldest C++11 GCC versions (4.9 and 5.0).

	2.3c - Feb 27, 2019
	---------
	Bugfix: Fixed assert failure at gstacks.cc:1171 (corrected with a return on gstacks.cc:1126)
	Bugfix: inadvertantly compiling out BAM support from process_radtags due to the removal of the HAVE_BAM
                config option, which occurred when we moved the BAM library internally to Stacks.
	Bugfix: corrected infinite loop in populations when --write-single-snp and -r were enabled.
        Bugfix: correct missing comment marker in population's FASTA exports and missing ']' character in FASTA comments.

	2.3d - Feb 28, 2019
	---------
	Bugfix: the snps_per_loc_postfilters distribution in the populations.log.distrib file was slightly off due to counting
                the number of SNPs at loci where despite SNP objects present at the locus, all sites were fixed in the
	        focal populations.
	Bugfix: Some haplotypes could pass through the filter, even after particular SNPs were filtered from them.
	Bufgix: Corrected the samples per locus and absent-samples per locus distributions from populations. 

	2.3e - Mar 20, 2019
	---------
	Bugfix: the --write-random-snp flag was causing an infinite loop in populations.

Stacks 2.2 - Aug 22, 2018
--------------------------
    Feature: Added the --bestrad flag to process_radtags. When used it will look for reads that need transposed
             before they are processed.
    Feature: gstacks: New option --max-debrujin-reads to control the construction of the de Bruijn graph; replaces
             --min-kmer-freq which is now deprecated
    Bugfix: Fixed a breaking circumstantial segmentation fault in populations
    Bugfix: Added run number to output FASTQ headers in process_radtags to make sure read IDs are always unique.

Stacks 2.1 - June 25, 2018
--------------------------
    Bugfix: Fixed a performance regression in sstacks. Recent changes in sstacks made it more likely to
            invoke the gapped algorithm to match to the catalog. In some cases, matches to the catalog
            would be marked as ambiguous alignments and dropped from the next stages of analysis due to
            differences in CIGAR strings from the gapped alignment.
    Feature: ustacks Changed --high_cov_thres default value from 2.0 to 3.0.
    Feature: gstacks: Changed --min-kmer-freq default value from 0.05 to 0.01.
    Feature: Added further checks on zlib calls.

Stacks 2.0 - Apr 23, 2018
-------------------------
    Feature: modified cstacks and ustacks gapped alignment algorithms to always align the two stacks/alleles
             with the most k-mers in common, removed the previous minimum k-mer limit.
    Feature: modified cstacks so that when a sample locus matches two or more catalog loci, those catalog
             loci are combined, or rolled-up, reducing undermerged loci that generate excess homozygote calls.
    Feature: modified tsv2bam so that when two loci from the same sample match the same catalog locus, those
             loci are combined.
    Bugfix: corrected BbvCI restriction enzyme to add the missing negative strand sequence.
    Bugfix: corrected the catalog writing routines for unzipped output files to include missing column.
    Bugfix: populations: Fixed the filtering of monomorphic loci.
    Bugfix: populations: Now preserving sample ordering in all outputs.
    Bugfix: Fixed ICPC compilation.

    2.0b - May 1, 2018
    Feature: gstacks: Removed the assertion that the first basepair of each locus should
             be part of a cutsite (now a warning).
    Feature: gstacks: The reported effective coverage is now a more realistic weighted mean.
    Bugfix: populations: Fixed STRUCTURE output being corrupted for some unordereds population maps.

Stacks 2.0 Beta 10 - Apr 10, 2018
---------------------------------
    Feature: Improved gapped alignment for secondary reads in ustacks.
    Feature: Improved populations performance.
    Feature: Added enzymes Cac8I, MslI.
    Feature: Made population maps more tolerant to spurious extra spaces and lines.
    Feature: populations: VCF output: changed the format of the catalog locus field and made the column 1-based.
    Feature: gstacks: increased haplotyping rates by adding a filtering of spurious SNPs step.
    Bugfix:  Fixed populations dramatric slow-down on datasets with more than several hundred samples.
    Bugfix:  Restored the NS/locus distribution in populations's distributions log.
    Bugfix:  Fixed the populations --radpainter export.
    Bugfix:  stacks-dist-extract: Fixed OSX compatibility.
    Bugfix:  Fixed breaking bug in populations --in-vcf mode filtering statistics.

Stacks 2.0 Beta 9 - Mar 12, 2018
--------------------------------
    Feature: Cleaned up tags/snps/alleles/matches files. We removed the batch ID from ustacks and cstacks
             output, and the deprecated log likelihood fields from ustacks and cstacks. We also removed
             the chromosome/bp/strand fields as they are no longer used in these files.
    Feature: Renamed gstacks output files that represent the new components of the catalog:
               gstacks.fa.gz => catalog.fa.gz; gstacks.vcf.gz => catalog.calls
    Feature: Removed read length restrictions from ustacks/cstacks/sstacks core, reads/loci can vary in
             length throughout the pipeline.
    Feature: Reimplemented PLINK export format for the populations program.
    Bugfix: Updated to HTSLib 1.7; changed to a custom build system that will work with the Stacks build
            system.
    Bugfix: Made gapped alignments mandatory in ustacks, cstacks, and sstacks. Added check for frameshift
            at 3' end of the read -- if found, a match is deferred to the gapped aligner.

Stacks 2.0 Beta 8 - Feb 03, 2018
--------------------------------
    Feature: populations: Now calculated deviation from Hardy-Weinberg equilibrium at the SNP level
             (using an exact test), and at the haplotype level (using Guo+Thompson's MCMC algorithm).
    Feature: populations: Added an export type for FineRADStructure.
    Feature: populations: Added the GQ/GL fields in the VCF SNPs output.
    Feature: gstacks: Made the default behavior regarding paired-end reads more logical (in
             reference-based mode --paired has been replaced with --unpaired).
    Feature: gstacks: Added details about samples and coverages to the log outputs.
    Feautre: Added enzymes NspI, BbvCI, fixed BfuCI.
    Bugfix:  corrected a major performance bottleneck in populations when smoothing population statistics
             across the genome.
    Bugfix:  populations: The VCF output now preserves the input sample order.
    Bugfix:  gstacks: Fixed the handling of a rare special case in the PCR duplicates code.
    Bugfix:  gstacks: Fixed 100% being added to all per-thread timings.

Stacks 2.0 Beta 7 - Dec 29, 2017
--------------------------------
    Feature: gstacks: Added an option to remove PCR duplicates based on insert
             size (--rm-pcr-duplicates, plus the related --rm-unpaired-reads).
    Feature: populations: Added a haplotype Genepop export.
    Feature: populations: improved the help; changed the output names for SNP
             files to 'populations.snps.EXT'; added option --no_hap_exports.
    Feature: gstacks and populations: Clarified the logs; moved distributions
             to a separate '.xlog' file and added script stacks-xlog-extract.
    Feature: gstacks: Tweaked the help/interface; especially, replaced --spacer
             with --suffix (for BAM directory input).
    Feature: Added enzymes BfuI and HinP1.
    Feature: Added option --inline_null to clone_filter.
    Bugfix: gstacks: Fixed a typo preventing the paired reads from being merged.
    Bugfix: populations: Fixed a segfault that occurred with some large datasets.
    Bugfix: Made VCF outputs more standard compliant.
    Bugfix: populations: Repaired --fasta_samples and --fasta_samples_raw.
    Bugfix: populations: Fixed population aborting at the end of the run
            when an export option was specified multiple times.
    Bugfix: gstacks: Adjusted progression report for catalog asymmetry.
    Bugfix: Fixed installation of stacks-integrate-alignments on MacOS.

Stacks 2.0 Beta 6 - Dec 02, 2017
--------------------------------
    Feature: Implmented the VCF haplotypes output.
    Bugfix:  Corrected asset failure in populations when exporting data for genepop or structure output.

Stacks 2.0 Beta 5 - Nov 27, 2017
--------------------------------
    Feature: Reimplemented structure, phylip, and phylip_var exports.
    Bugfix: Tightened up the overlap algorithm to require 80% of overlapping sequence to be
            aligned and of the aligned sequence, 80% must be identities.
    Bugfix: Fixed segfault in gstacks when compiled with CLANG on OS X.
    Bugfix: gstacks: Fixed how misphasings are reported.

Stacks 2.0 Beta 4 - Nov 07, 2017
--------------------------------
    Bugfix: Continued improving overlap algorithm to join SE and PE contigs.
    Bugfix: Improved build system to handle new timing functions in gstacks.

Stacks 2.0 Beta 3 - Nov 01, 2017
--------------------------------
    Feature: Added output to populations describing mean PE contig size and mean number of
             genotyped sites per locus, which reflects the current filtering paramters.
    Feature: Improved the output of gstacks and populations.
    Feature: Added script `stacks-integrate-alignments`.
    Bugfix: made further improvements to the single-end/paired-end locus overlapping algorithm.
    Bugfix: fixed all depths being null in populations' VCF output.
    Bugfix: Numerically tweaked the marukilow model to remove a limit case.

Stacks 2.0 Beta 2 - Oct 19, 2017
--------------------------------
    Feature: gstacks: Made it possible to read from multiple BAM files at the same time; modified the
             interface accordingly.
    Feature: gstacks: Parallelized the reference-based mode.
    Feature: gstacks: Added various statistics & improvements to the log output.
    Feature: gstacks: Improved how the forward & paired-end reads are merged (in denovo mode; no more trimming).
    Feature: populations: Added code to calculate the overlap between RAD loci when a reference is available.
    Feature: populations: Added VCF ouput (--vcf).
    Feature: Updated the denovo_map.pl and ref_map.pl wrappers, samples must now be specified using --samples and --popmap.
    Bugfix: Fixed three memory leaks in populations; improved reference-aligned batch logic.
    Bugfix: Improved overlapping code in gstacks to merge more single and paired-end contigs together.
    Bugfix: Now compiles on Apple OS X.
    Bugfix: Fixed a bug that skewed the fixed-site (no-SNP) likelihood in the marukilow model.

Stacks 2.0 Beta 1 - Oct 09, 2017
--------------------------------
    Feature: Paired-end sequencing data can be utilized fully. In particular, when the shearing-based
             protocol is used, the software will assemble a local contig from the paired reads across
             the population, possibly overlap it with the forward-reads region, then align all reads to the
             assembled contig. This new approach also fully supports double-digest protocols.
    Feature: Haplotype calling and diploidy-violation dectection now rely on a novel, more powerful algorithm.
    Feature: SNP and genotype-calling now uses the diploid models of Maruki and Lynch (2017).
    Feature: The rxstacks program has been replaced with the gstacks program, and there is no need to re-run
	     some of the earlier steps of the pipeline anymore.
    Feature: The memory footprint of the populations program has been considerably reduced and can be scaled
	     for any size data set.
    Feature: The reference-based pipeline has been simplified, and now only comprises two steps: gstacks and populations.
    Feature: Added --null_inline mode to clone_filter (and process_radtags) for previously unseen type
             of oligo combination.

Stacks 1.48 - Nov 20, 2017
---------------------------
    Feature: Added HinP1I restriction enzyme.
    Feature: Added --null_inline mode to clone_filter (and process_radtags) for previously unseen type
             of oligo combination.

Stacks 1.47 - Sept 06, 2017
---------------------------
    Feature: Improved populations's fasta output options (especially,
             added a option to export locus consensus sequences).
    Feature: denovo_map.pl and red_map.pl now stop if a component
             of the pipeline fails.
    Feature: Improved the output of denovo_map.pl and ref_map.pl.
    Bugfix: Added a format check in Fasta/GzFasta to avoid a potential
            segfault when working on FastQ files.
    Bugfix: Fixed a bug in count_fixed_catalog_snps.py that could cause
            overwrites when working with uncompressed files.

Stacks 1.46 - Apr 17, 2017
--------------------------
    Feature: Added HaeIII enzyme.
    Bugfix: Corrected memory leaks in rxstacks.
    Bugfix: Corrected non-functioning --min_mapq parameter for pstacks.
    Bugfix: Corrected segfault when combining a VCF input file to populations,
            with genomic output and masking a restriction enzyme.

Stacks 1.45 - Feb 24, 2017
--------------------------
    Feature: Tweaked the interfaces of most programs:
             * cstacks and sstacks now accept a population map as input.
             * process_radtags will now reuse the input directory name in its log file name.
             * Reworked pstacks output.
             * Batch ID now defaults to 1 in cstacks, and sstacks and other will try to guess
               it from the contents of the given directory/catalog path.
             * pstacks/ustacks/process_radtags will now try to guess file formats.
             * Default (fallback) format in process_radtags/process_shortread is now gzfastq.
             * pstacks: Substituted --max_clipped to --min_aln_pct.
             * ustacks -r has become the default; --keep-high-cov reverses it.
             * cstacks now checks for sample ID unicity.
             * Updated help messages.
    Feature: populations now logs the 'number of SNPs per locus' distribution.
    Feature: Added mapping quality filter in pstacks (--min_mapq).
    Feature: Added enzyme ApaLI.
    Bugfix: populations: Corrected a VCF-related segfault (current use of VCF's GL field was
            improper and was removed).
    Bugfix: rxstacks: Corrected a bug that affected locus likelihood medians.
    Bugfix: pstacks/ustacks: Corrected a bug that affected coverage standard deviations.
    Bugfix: populations: Fixed parsing of option --sigma.
    Bugfix: Fixed process_radtags writing fasta (instead of fastq) discard files when input
            files were gzfastq.
    Bugfix: kernel smoothing was not working correctly for Fis values (values were too negative).
    Bugfix: fixed a regression for gapped alignments in cstacks that was causing a buffer overflow.

Stacks 1.44 - Oct 11, 2016
--------------------------
	Bugfix: corrected an error in pstacks where '=' and 'X' symbols were not recognized properly in SAM/BAM
	CIGAR strings.
	Bugfix: corrected some typos in pstacks/populations help output.

Stacks 1.43 - Oct 05, 2016
--------------------------
	Feature: added alignment controls to pstacks, allowing the program to discard secondary alignments
	and to discard alignments where a significant portion of the read was not aligned (soft-masked).
	Bugfix: corrected a very small memory leak in the gapped alignment code, found by Valgrind.
	Feature: updated configure test to check if compiler can handle c++11 standard.
	Bugfix: rxstacks was not generating model files.
	Bugfix: corrected an uncaught exception in cstacks when processing gapped alignments. In some cases when a
	multiple alignment had to be recomputed the initial CIGAR string was not parsed properly leading to the
	catalog and query sequences coming out of sync in their length (which could throw the exception).
	Feature: reduced memory usage in ustacks and pstacks by not retaining all reads from a collapsed locus.
	Bugfix: corrected -V option for populations, which was causing a crash (although --in_vcf worked).

Stacks 1.42 - Aug 05, 2016
--------------------------
	Feature: Added Csp6I restriction enzyme.
	Feature: populations program is now able to calculate populations statistics using arbitrary VCF files
	as input.
	Feature: Upgraded to the latest release of HTSLib (1.3.1) for reading BAM files. Embedded the library
	in the Stacks distribution to remove previous libbam dependency.
	Feature: Added an output directory option to 'populations' (--out_path).
	Feature: Added restriction enzymes BsaHI, HpaII, NcoI; corrected NdeI.
	Bugfix: Made the VCF output by 'populations' more standard-compliant.
	Bugfix: Some output files included 0-based genomic coordinates, changed them to 1-based.
	Bugfix: Replaced populations IDs with populations names in 'populations' output.
	Bugfix: Corrected a bug affecting clone_filter when input was non-gzipped paired-end data.

Stacks 1.41 - June 22, 2016
---------------------------
	Bugfix: the kernel-smoothing procedure in populations (used for Fst, Pi, heterozygosity etc. smoothing)
	was not functioning at sizes larger than the default size. A bug was creating incorrect weights for the
	smoothing operation when the sliding window size was set to a large value causing the smoothing
	window to have a maximum size after which increasing the size did not change the smoothing.
	Bugfix: cstacks was reporting gapped alignments even when --gapped was not enabled. This affected
	a small number of (mostly) confounded catalog loci.
	Feature: Added the Csp6I restriction enzyme.

Stacks 1.40 - May 04, 2016
--------------------------
	Feature: Changed process_radtags and process_shortreads to print FASTQ/FASTA headers using
	"/1" and "/2" to represent the read number, instead of "_1" and "_2".
	Bugfix: fixed a regression where allele depths were not being loaded due to the use of the new
	*.models.tsv file. This file lacks the raw reads and therefore we can't count the raw stack depth
	when running sstacks.
	Bugfix: cstacks was calling errant SNPs in loci with a sample containing one gapped locus and
	one ungapped locus matching the same catalog locus.

Stacks 1.39 - April 23, 2016
----------------------------
	Bugfix: rxstacks was not adjusting reads/SNPs to account for alignment gaps. There was also an
	bug in reading the input files.
	Bugfix: denovo_map.pl and ref_map.pl were not processing parents/progeny properly.

Stacks 1.38 - April 18, 2016
----------------------------
	Feature: denovo_map.pl and ref_map.pl now print depth of coverage for each sample. The ustacks
	program now prints depth of coverage after each algorithm stage to see how each stage improves
	(or not) the depth of coverage.
	Feature: complete refactoring of denovo_map.pl and ref_map.pl. Separated computation from
	SQL loading. Added auto creation/deletion of database. Enabled samples to be read from population
	map instead of specifying them on the command line.
	Feature: added Needleman–Wunsch algorithm to ustacks, cstacks, sstacks to provide for gapped
	alignments. Includes --max_gaps and --min_aln_len parameters to contain crazy
	alignments. sstacks now includes a CIGAR string describing the alignment to the catalog.
	Feature: optimized ustacks for a 33% decrease in run time.
	Feature: added new file, sample_X.models.tsv.gz, produced by ustacks and pstacks. Contains a subset
	of the information in the sample_X.tags.tsv.gz file, allows for data to be loaded much faster in the
	later stages of the pipeline, greatly speeding up run times.
	Bugfix: added code to prevent populations from improperly reading SNP positions past the length of
	a particluar locus (that is shorter than the catalog locus).
	Bugfix: corrected bug in process_radtags when using inline barcodes on paired-end reads. The paired-
	end reads were not being truncated uniformly.
	Bugfix: corrected bug in populations where if enough empty files were fed into the program
	it could place files in the wrong population or segfault.
	Bugfix: corrected PHP files for exporting to include LnL filter.
	Bugfix: corrected mappable markers filter in web interface.

Stacks 1.37 - Feb 24, 2016
--------------------------
	Feature: converted PHP database code from MDB2 to MySQLi. MDB2 is no longer a
	prerequisite for installing Stacks.

Stacks 1.36 - Feb 18, 2016
--------------------------
	Feature: Added the BfaI, BspDI, AseI, and AciI restriction enzymes to process_radtags.
	Feature: Changed the way denovo_map.pl and ref_map.pl run sstacks. It is now set to run
	sstacks once for all samples, instead of one time per sample. Should provide a significant
	speed-up.
	Bugfix: corrected error in pstacks when handling long reads with complex SAM/BAM alignments.
	Bugfix: fixed memory leak in sstacks when more than one sample file was specified.
	Bugfix: corrected error in clone_filter causing it to fail when processing gzipped data
	without a random oligo attached.
	Bugfix: corrected error when reading gzipped FASTA files and the last sequence of the file
	was improperly doubled in length.

Stacks 1.35 - Sept 09, 2015
---------------------------
	Feature: Added --retain_header flag to process_radtags/process_shortreads which will keep
	the unmodified FASTQ header in the output. This allows clone_filter/process_radtags/
	process_shortreads to be run in different sequences and more than one time.
	Feature: Added --treemix to the populations program, allowing SNPs to be output in
	TreeMix format.
	Feature: Added --phylip_var_all to the populations program. This option outputs the full
	sequence from each variable locus, encoding polymorphisms using IUPAC notation.
	  -This option will also output a file containing the coordinates of each RAD locus so they
	  can be input to phylogenetic software (such as RAxML) to partition each RAD locus out
	  and then build the phylogenetic tree independently for each partitioned locus.
	Feature: Added the AgeI restriction enzyme.
	Feature: refactored clone_filter to handle random oligo sequences used as inline/indexed
	barcodes to identify and discard PCR duplicates.
	Bugfix: added code to process_radtags/process_shortreads to handle cases when data writes
	fail due to a filled disk or other error conditions.
	Bugfix: kmer_filter was not handling gzipped FASTQ files properly when filtering rare kmers.

Stacks 1.34 - July 26, 2015
---------------------------
	Bugfix: fixed phylip output to again include nucleotides from subsets of the full set
	of populations.
	Bugfix: private alleles were being associated to the incorrect population at a particular
	locus (the counts and summary statistics of private alleles were not affected).

Stacks 1.33 - July 22, 2015
---------------------------
	Bugfix: Corrected the second-stage filtering of the populations program to properly
	respect the -p flag.
	Bugfix: Corrected the display of individual samples in the web interface (tags.php file).

Stacks 1.32 - June 18, 2015
---------------------------
	Bugfix: Updated the Phylip output to reflect the changed meaning of 'fixed' as
	determined in the PopSum::tally() function.

Stacks 1.31 - June 17, 2015
---------------------------
	Bugfix: site-level filtering in the populations program was not working correctly
	when dealing with sites that were fixed within populations but variable among
	populations. The code in the PopSum::tally() function was not correctly identifying
	sites as not fixed in these cases causing them to be incorrectly filtered out.
	Bugfix: --write_random_snp was causing a segfault in the populations program in
	some cases.
	Feature: changed the default setting for the -n option of cstacks (number of fixed
	differences allowed between loci) to 1 (at the request of
	Josie Paris <J.R.Paris@exeter.ac.uk>).
	Bugfix: made some tweaks to improve layout in the web interface.
 	Bugfix: single-end reads, with paired barcodes (inline/index) were not being handled
	properly, resulting in a segfault.
	Bugfix: process_radtags was allowing a non-null barcode type to be specified without
	specifying a barcode file, which caused a segfault.
	Feature: exposed kmer length setting in ustacks and cstacks. This allows the kmer
	length used for sequence matching to be set manually. While this can result in some
	missed matches (there is a trade off between kmer length and sequence length when
	searching for matches between the two) it also allows the algorithm to run at faster
	speeds.
	Feature: Changed default database engine type to be excplicitly MyISAM. Previously
	Stacks just used the default which at one time was MyISAM but has recently changed in
	many systems to be INNODB. Using MyISAM should provide much faster imports of data
	and ultimately use less disk space (as the space is reclaimed when databases are
	deleted).

Stacks 1.30 - May 07, 2015
--------------------------
	Feature: sstacks can now accept multiple sample files at a time, saving run
	time by only processing the catalog once.
	Feature: changed batch_X.sumstats.tsv file so the P/Q alleles are always
	presented in the same order in each local population (according to the
	overall frequency of the allele across all populations). This will sync results
	with the VCF exports but will sometimes cause the frequency of p in the local
	population to be less than 0.5 (up until now the frequency of p has always
	represented the most frequenct allele in the local population).
	Feature: added an maximum observed heterozygosity filter to populations program.
	Bugfix: Fis values in batch_X.sumstats_summary.tsv were incorrect (although raw
	values in batch_X.sumstats.tsv were correct).
	Bugfix: corrected the allele depth output in the VCF export to follow defacto
	standards used by other programs.
	Bugfix: in some cases loci were sneaking past the --write_single_snp directive in
	in populations (due to interactions with pruning out SNPs that fail the MAF filter).
	Feature: Updated the Stacks web interface. The web app is now almost 100% dynamic
	(parts of the page are draws on demand instead of fetching new, full pages from
	the server) using local javascript to draw the population view of genotypes, summary
	statistics, and the view of raw stacks. The web app uses asynchronous AJAX queries
	that trade data encoded in JSON to fetch the necessary data for dynamic display.
	Feature: added DdeI, RsaI, AluI restriction enzymes to process_radtags.
	Bugfix: sstacks could generate extra matching haplotpyes in a very small number of
	cases.

Stacks 1.29 - Mar 21, 2015
--------------------------
	Feature: added the --ordered_export option to the populations program. For the VCF,
	GenePop, and Structure exports, if this option is specified, only one copy of each
	SNP is exported in the case where one nucleotide position is covered by more than one
	RAD locus. Most useful for ddRAD data.
	Feature: VCF export now includes individual allele depths for each SNP call.
	Feature: improved the filtering logging code in populations, if the --verbose flag
	is specified, a reason is provided for each pruned site, or each removed locus.
	Bugfix: PHASE output was broken in the populations program. SNP pruning/filtering
	code did not update the catalog copies of the alleles after pruning which are needed
	by the PHASE output code.
	Bugfix: adjusted the filtering code in populations to not exclude fixed loci.
	Bugfix: removed extra tab from ID line for Structure export.
	Bugfix: fixed issue in genepop output that may have overfiltered some loci.
	Bugfix: fixed small problems with --write_single_snp/--write_random_snp in the
	populations program. Some polymorphic loci were erroneously being omitted.

Stacks 1.28 - Mar 06, 2015
--------------------------
	Feature: added a second barcode distance to process_radtags/process_shortreads.
	This allows you to specify two distances for recovering barcodes if you are using
	combinatorial barcodes (e.g. a 12bp barcode inline on the single-end read plus a
	6bp index). I have changed the meaning of the parameter from "distance between
	barcodes" to "number of allowed mismatches when correcting barcodes."
	   The --barcode_dist parameter is now --barcode_dist_1, and --barcode_dist_2
	   was added.
	Bugfix: the process_shortreads/process_radtags programs were trimming sequence
	as if an inline barcode was specified, even when it was an index barcode and no
	sequence should have been trimmed.
	Bugfix: the process_shortreads program was outputting FASTA even when FASTQ was
	requested due to not handling gzipped outputs properly.
	Bugfix: fixed segfault in populations that could occur when using a whitelist that
	contained loci that were being filtered out due to -p/-r constraints.

Stacks 1.27 - Feb 25, 2015
--------------------------
	Bugfix: the minor allele frequency filter and the proceny limit filter were not working
	properly in all cases with the other filters.
	Bugfix: barcode length (href->inline_bc_len) was not being correctly set for single-end,
	inline line barcodes of variable length.

Stacks 1.26 - Feb 23, 2015
--------------------------
	Bugfix: if you are running non-compressed data, then version 1.25 broke the parsing code.
	If your data were zipped (or a BAM file) when it went through pstacks/ustacks, then
	there was no bug.
	Feature: refactored the filtering code in the populations program to add a second
	filtering step. In previous versions the -r (sample limit) and -p (population limit)
	were applied on the basis of the entire RAD locus. This could lead to situations
	where a RAD locus remained in the data set while one or more of the individual SNPs
	on that locus were missing data and were below the -r or -p limits. Now, the filters
	are applied to individual SNPs after the filters are applied to the RAD loci. This
	greatly affects the -r (sample limit) filter with more SNPs being pruned out, as well
	as the -a (minor allele frequency filter) such that all SNPs below the MAF are
	pruned fully from the data set and will not appear in any statistical results or
	downstream exports.
	Feature: added restriction enzyme kpnI.
	Feature: added code to check for the existence of the loci and SNPs provided in a
	whitelist.

Stacks 1.25 - Feb 17, 2015
--------------------------
	Feature: added support for unaligned BAM files for process_radtaags and
	process_shortreads. The two programs can now read paired data that is interleaved in
	a single file (which is required to support paired-end data in BAM format).
	Feature: Haplotypes can now be output in VCF format from the populations program using
	the --vcf_haplotypes option.
	Feature: added --fasta_strict option to populations program. Will output full sequence for
	each individual at each haplotype at each locus, but only for biologically plausible loci.
	It won't output loci with more than two haplotypes and will output single haplotypes twice,
	once per allele.
	Feature: Changed the sumstats/hapstats files to output a one-based genome base pair position
	so it matches other export formats.
	Bugfix: fixed problem with gzipped files where last line of file was not read properly
	causing the program to output an erroneous error message.
	Bugfix: The FASTA output from the populations program was reporting the internal value
	(zero-based index) for the basepair position of each read (the first nucleotide of the
	cutsite) causing an off-by-one error for all reads and reads on the negative strand had
	the coordinate for the cutsite end of the read (right-most end) reported instead of the
	standard left-most end.
	Bugfix: the log likelihood filter was not working properly in export_sql.pl, causing many
	genotypes to be excluded during export.
	Bugfix: process_radtags was not looking for the paired-end RAD cutsite in the proper location
	when dealing with double-digest, inline/index barcoded reads.
	Feature: added initial, internal support for merging and phasing loci that overlap at a
	restriction enzyme cut site.
	Feature: code now prints program version and generation date to all internal Stacks files.

Stacks 1.24 - Jan 07, 2015
--------------------------
	Feature: added restriction enzyme ecoRV.
	Bugfix: fixed segmentation fault in process_radtags/process_shortreads when resizing sequence
	and phred internal buffer sizes.

Stacks 1.23 - Dec 12, 2014
--------------------------
	Bugfix: Fixed a segfault bug in process_radtags where the process_barcode function returned
	prematurely when one barcode was correct and one was incorrect in paired cases.
	Bugfix: fixed compiler warnings when building with CLANG.

Stacks 1.22 - Dec 08, 2014
--------------------------
	Feature: process_radtags and process_shortreads now support variable barcode lengths. In
	process_radtags sequences will automatically be trimmed to keep stacks a uniform length
	with the variable barcode lengths.
	Feature: a filename can now be specified in the barcodes file for process_radtags and
	process_shortreads. When a filename is specified, process_radtags will write data to
	this filename instead of a filename made up of the barcode.
	Feature: process_radtags and process_shortreads will now output gzipped files if
	provided gzipped inputs or if requested using the '-y' option.
	Feature: Added SacI and BgIII restriction enzymes.
	Bugfix: Tightened up parsing of FASTQ ID field to prevent a buffer overrun (and subsequent
	segfault) in FASTQ headers that look like the Illumina format but are malformed.
	Bugfix: Fixed GenePop output of populations program as last locus on second line was missing
	commas if more than one SNP was present at that locus.
	Bugfix: -R option to retain unused reads was not being recognized by ustacks.
	Bugfix: changed populations to record program run parameters and execution time to log file.
	Bugfix: corrected Makefile.am to include Sparsehash compile flags for process_radtags.
	Bugfix: corrected load_radtags.pl so as not to try and load the population ID as a
	number to the samples table (and instead as a string).

Stacks 1.21 - Oct 02, 2014
--------------------------
	Feature: Added the XbaI, BstYI, and XhoI restriction enzymes.
 	Feature: Added ability to specify column position in whitelist along with locus ID in
	populations program. This allows for specific SNPs within specific loci to be processed.
	Feature: In populations program, changed implementation of --write_single_snp to create
	an internal whitelist from the first SNP in each catalog locus. Added a new command
	line option, --write_random_snp to select a single, random SNP per RAD locus using the
	same internal mechanism.
	Feature: Added HZAR, Hybrid Zone Analysis in R output to populations program.
	Bugfix: Added code in populations program to handle cases where a haplotype contains one
	or more uncalled bases (Ns). These haplotypes are now excluded from haplotype-based
	statistical calculations.
	Bugfix: In Phi_st/ct/sc calculations of populations program, total population count was
	not adjusted downward when one of the populations dropped out of the analysis at a
	particular locus in the all-populations, haplotype-based AMOVA calculation (batch_X.phistats.tsv).
	Bugfix: "All positions" Fis measure in batch_X.sumstats_summary.tsv file too negative due
	to internal logic error.
	Bugfix: updated queries in index_radtags.pl to account for new 'type' variable in SNPs tables.

Stacks 1.20 - Jul 29, 2014
--------------------------
	Synced corrections module branch with main Stacks branch.
	***
	The internal formats of the *.tags.tsv, *.snps.tsv, and *.matches.tsv files have changed
	and therefore version 1.20 programs cannot be used on earlier generated data sets. However,
	the convert_stacks.pl script is included in this release to convert an older data set into
	the new formats.
	***
	Feature: Implemented new haplotype trimming algorithm for rxstacks.
	Feature: new script, convert_stacks.pl, to convert previous Stacks files to new format.
	Feature: Modified VCF output to include likelihood values from heterozygous and homozygous
	SNP model calls.
	Feature: added log likelihood filter to genotypes and populations programs and to web interface.
	Feature: Added SpeI restriction enzyme to process_radtags.
	Feature: Modified Beagle output formats in populations program to be population-specific and
	not to include monomorphic nucleotide positions.

Stacks 1.19 - Apr 23, 2014
--------------------------
	Feature: the populations program now calculates Fst' and D_est on haplotypes between all pairwise
	populations. Our implementations are based on:
	    Bird, Karl, Smouse & Toonen. (2011) Detecting and measuring genetic differentiation.
	    D_est: Jost. (2008) Gst and its relatives do not measure differentiation.
	    Fst': based on modifying the AMOVA implementation from Excoffier, Smouse, & Quattro (1992).
	Feature: we have refactored the populations program to use a common framework for kernel smoothing
	and bootstrapping. This has allowed us to add smoothing and bootstrapping to all statistics calculated
	by the populations program: pi, Fis, Fst, Fst', D_est, Phi_st, Phi_ct, Phi_sc, Haplotype diversity,
	gene diversity.
	Feature: we have implemented fine-grained control of bootstrapping by providing flags to turn on
	bootstrapping for each group of population statistics, as well as providing a bootstrapping whitelist
	allowing only certain loci to be included in the bootstrapping calculations.

Stacks 1.18 - Apr 04, 2014
--------------------------
	Feature: we now use chi squared segregation ratios to detect missing alleles in parental mapping markers.
	in F1 crosses (CP map type). We can now map ab/a- and -a/ab as ab/--, and --/ab markers; we can map
	ab/c- and -c/ab markers as ab/cd markers; we can map aa/b- and -a/bb markers as ab/-- and --/ab markers.
	Feature: in F1 crosses we are now mapping ab/cc and cc/ab markers as ab/-- and --/ab markers.
	Feature: reworked genetic map display of web interface. Included chisq p-value from segregation distortion
	test as a filter.
	Feature: implemented measure of segregation distortion in genotypes program based on chi square test of
	genotype counts. Removed deprecated measure of F, inbreeding coefficient, replaced it with segregation
	distortion.
	Bugfix: corrected calling of markers in genotypes program. When a whitelist with a small number of markers
	is specified, some of the parental IDs could be missed, causing markers not to be called and hence dropped
	from the analysis.
	Bugfix: changed genotype mappings for generic map types to make certain non-biologically plausible genotype
	combinations illegal.
	Bugfix: fixed compilation issues when using Google's SparseHash (thanks to khuck@cs.uoregon.edu for the patch).

Stacks 1.17 - Mar 26, 2014
--------------------------
	Bugfix: Added #ifdefs to deal with missing functions in older versions of zlib.

Stacks 1.16 - Mar 25, 2014
--------------------------
	Feature: added haplotype counts for each population and locus to the batch_X.hapstats.tsv file.
	Feature: haplotype F statistics are now calculated for the whole set of populations (one analysis
	of variance calculation for all populations), and also as a set of pairwise calculations to mirror
	the existing Fst calculations.
	Bugfix: fixed small bug in calculation of MSD(Total) component of Phi_st (haplotype F statistics).
	Bugfix: fixed bug in parsing of populations maps when using strings for population identifiers.
	Bugfix: kernel-smoothing not correct for haplotype/gene diversity.

Stacks 1.15 - Mar 15, 2014
--------------------------
	Bugfix: fix various bugs related to gzip support.

Stacks 1.14 - Mar 14, 2014
--------------------------
	Feature: Stacks files are now kept in gzipped format if FASTQ data is fed into pipeline gzipped or as a BAM.
	Bugfix: fixed some compile bugs on OSX Mavericks.

Stacks 1.13 - Feb 24, 2014
--------------------------
	Feature: We have implemented the first set of haplotype-level population genetics statistics. Specifically,
	we are now calculating gene diversity and haplotype diversity (pi) for each locus, as well as F statistics
	for haplotypes including, Phi_st, Phi_ct, and Phi_sc, which are calculated using Analysis of Molecular
	Variance (AMOVA):
	    Excoffier, Smouse, & Quattro, (1992). Analysis of molecular variance inferred from metric distances
	    among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics.
	Data can be analyzed as populations of individuals (the previous default) and now using populations
	of individuals, and groups of populations.
	Feature: If a reference genome is available, haplotype F statistics can also be kernel-smoothed.
	Feature: populations in the population map can now be specified as text strings or numbers. Groups
	of populations can now be specified by adding a third column to the population map for each individual
	and listing the group they belong to (again as a text string or number).
	Bugfix: allow batch IDs of 0 in populations and genotypes.
	Bugfix: in populations, changed VCF output to be ordered by basepair.
	Bugfix: in populations, change value of expected homozygosity to be set to 1 - expected heterozygosity
	instead of 1 - Pi. Pi (computed as [1 - ((p choose 2) + (q choose 2) / (n choose 2))] and expected
	heterozygosity (2pq) can produce sligthly different estimates resulting in exp het + exp hom != 1.

Stacks 1.12 - Jan 21, 2014
--------------------------
	Bugfix: accidentally broke gzipped FASTQ support through a typo in gzFastq.h.

Stacks 1.11 - Jan 09, 2014
--------------------------
	Feature: changed build to work properly with g++ and clang, which is the native compiler on
	Apple's OS X.
	Feature: Added NheI restriction enzyme.
	Bugfix: changed logging in denovo_map.pl/ref_map.pl to write outputs from Stacks programs continuously
	instead of waiting until the program completed to write output to log file.
	Bugfix: corrected parsing of population map for gzipped input files for denovo_map.pl.

Stacks 1.10 - Dec 10, 2013
--------------------------
	Feature: Added phased output for PHASE and Beagle. The phased output writes multiple SNPs
	in a single RAD locus as an already phased haplotype, leaving PHASE and Beagle to only phase
	between these haplotypes, instead of having to re-phase SNPs from within the same RAD site.
	Bugfix: corrected the SNP genotype output for Beagle.
	Bugfix: Corrected PHP warnings; enabled scrolling in catalog.php for iframes.
	Bugfix: allele percentages from ustacks were off since ustacks was changed to load/unload
	read IDs from disk (Stacks 0.99995). Only the calculation of the percentages was affected,
	not the underlying algorithms.

Stacks 1.09 - Oct 30, 2013
--------------------------
	Feature: added export support for F2 and backcross map types for Onemap to genotypes.
	Feature: added EaeI, ClaI, and TaqI restriction enzymes to process_radtags.
	Feature: changed populations bootstrap to use AMOVA Fst.
	Feature: added bootstrap whitelist to populations, so users can restrict the loci that
	are bootstrapped to a particular set (e.g. on a single chromosome).
	Bugfix: modified PHASE output so that SNPs are ordered properly. Previously, although
	RAD loci are ordered properly, some individual SNPs between RAD loci could still be output
	out of order.
	Bugfix: corrected onemap CP output so that B3.7 markers are output as "ab", not "2ab".

Stacks 1.10.Beta1 - Sept 30, 2013
---------------------------------
	Feature: completed implementation of rxstacks.
	Bugfix: when merging a homozygous locus into the catalog, if homozygous allele conflicted
	with existing catalog SNP alleles, new allele was not added to SNP object (but was added to
	the allele list).
	Bugfix: found small memory leak in cstacks - old SNP objects were not being freed when new
	SNPs were merged into the catalog.
	Bugfix: empty alleles were being output to the batch_X.catalog.alleles file by cstacks. Did
	not affect the function of the program.

Stacks 1.08 - Sept 24, 2013
---------------------------
	Feature: added a FASTA output to populations to output the full locus sequence
	for each allele at each sample locus, applying any filters or whitelists supplied to
	populations.

Stacks 1.07 - Sept 23, 2013
---------------------------
	Bugfix: updated process_radtags to drop reads shorter than length
	limit when read trimming turned on.
	Bugfix: corrected build failures on Mac OS X due to Samtools' bam.h header conflicting
	with Stacks' Bam.h header when building on OS X's case insensitive file system.
	Feature: changed process_radtags to drop reads already shorter than limit if sequence
	truncation turned on. You can also specify the read length limit to drop reads if your
	data have already been trimmed.
	Bugfix: Updated VCF ouput, missing genotypes now reported as "./." instead of "."
	Bugfix: Updated VCF ouput, alleles reported on the negative strand are now complemented
	so their positive strand conterparts are reported and will align aginst a reference genome.
	Bugfix: Updated VCF ouput, "reference allele" is now always reported as most frequent allele.

Stacks 1.06 - August 28, 2013
-----------------------------
	Bugfix: Illumina FASTQ header specifying read pair could override internal enumeration
	of read pair if paired-end data was fed in as a single-end file.
	Bugfix: corrected locus starting base in reference-aligned data.
	Feature: refactored sort_read_pairs.pl to process input files one at a time, without retaining
	them in memory. The program should now be able to handle an arbitrary number of samples.
	Feature: sort_read_pairs.pl can now read gzipped files directly.

Stacks 1.05 - August 17, 2013
-----------------------------
	Bugfix: adapter filtering code in process_radtags/process_shortreads bit rotted and was not
	properly functioning. Switching from deprecated hash function to TR1 hash broke the expected
	hashing	behavior for char *.
	Bugfix: modified process_radtags/process_shortreads to handle single adapters when processing
	paired-end data (previously you had to specify two adapters for paired data).
	Bugfix: corrected barcode-specific counters in process_radtags/process_shortreads. Overall counts
	were correct but counts for barcodes were off due to shuffling of code that happened with support
	of combinatorial barcodes.

Stacks 1.04 - July 25, 2013
---------------------------
	Bugfix: process_radtags was not properly handling index_index and inline_inline barcode types.
	Bugfix: the hindIII restriction enzyme sequence was incorrectly specified in renz.h.
	Bugfix: ustacks wasn't properly removing file suffix when gzip files are processed.

Stacks 1.03 - June 28, 2013
---------------------------
	Bugfix: non-barcoded data were not being handled properly by process_radtags/process_shortreads.

Stacks 1.02 - June 24, 2013
---------------------------
	Bugfix: single-end barcode, double-digested data were not being handled properly by
	process_radtags causing a crash.
	Feature: added support for PLINK and Beagle output files from the populations program.
	Feature: Modified the minor allele frequency (MAF) filter to remove polymorphic nucleotide
	SNPs from Stacks output on a per-population basis. So, if a second allele is present
	at a frequency below the MAF, that nucleotide site is not output (although other sites
	at the same RAD locus could still be output).
	Bugfix: Tri-allelic loci were being output into the STRUCTURE, GENEPOP and PHASE
	output (but not in sumstats or Fst).

Stacks 1.01 - June 07, 2013
---------------------------
	Bugfix: an off-by-one error was preventing haplotypes from being verified by sstacks
	if a SNP occurred in the last position of the read. This could cause tags to fail
	to match to the catalog if there is a SNP in the final position.

Stacks 1.0 - June 06, 2013
--------------------------
	Feature: added XbaI and BamHI restriction enzymes to process_radtags.
	Feature: added code to output genotypes in PHASE/fastPHASE format.
	Feature: extended combinatorial barcodes support so one can process single-end data
	that contains both an inline and indexed barcode.
	Feature: added command line option and supporting code to cstacks to allow samples
	to be added to an existing catalog.
	Feature: refactored command line handling in denovo_map.pl and ref_map.pl to be much
	more flexible. Arbitrary command line options can now be passed to particular pipeline
	programs using the -X flag.
	Feature: for genetic maps, catalog may now be constructed out of mulitple parents,
	genotypes is smart enough to cross check the parents used to construct the catalog
	against those submitted to genotypes for producing a map. Will allow for a single
	catalog to be used across a series of crosses so all maps share the same catalog IDs.
	Feature: added option to genotypes to import manual corrections exported from Stacks
	SQL database.
	Feature: added --log_fst_comp option to populations to log components of the Fst
	calculations to a file for debugging / testing purposes.
	Bugfix: corrected handling of files in kmer_filter. Adding support for gzipped files
	broke file handling in some cases.

Stacks 0.999991 - May 14, 2013
------------------------------
	Feature: changed populations to use AMOVA Fst for batch_1.fst_summary.tsv file. Previously
	it used the Binomial Fst.
	Bugfix: If --write_single_snp not specified, Structure output was not naming loci properly (it
	was naming each SNP from the same RAD locus using the same ID, instead of differentiating each
	SNP in each RAD locus).
	Feature: Added Sau3AI and SexAI restriction enzymes. Fixed bug in specificaion of MseI, MspI
	enzymes.
	Bugfix: changed VCF and Fst code in populations to output SNPs from reads aligned to the
	negative strand on a reference genome correctly.

Stacks 0.99999 - May 06, 2013
-----------------------------
	Bugfix: process_shortreads/process_radtags not working with non-barcoded data.

Stacks 0.99998 - May 01, 2013
-----------------------------
	Feature: Added option to sort_read_pairs.pl to output FASTQ if desired.
	Bugfix: make sort_read_pairs.pl understand new file naming scheme.
	Feature: added mseI, mspI restriction enzymes to process_radtags.
	Bugfix: corrected sphI cutsite sequence in process_radtags.
	Bugfix: stopped "uninitialized value" errors in export_sql.pl when marker type is
	undefined for a particular map.

Stacks 0.99997 - April 01, 2013
-------------------------------
	Bugfix: paired barcode could become uninitialized on second pair of files in
	process_radtags/process_shortreads causing all barcodes to mismatch. Made Read
	class explicitly initialize everything.

Stacks 0.99996 - March 24, 2013
-------------------------------
	Feature: major overhaul of the process_radtags / process_shortreads programs to support
	combinatorial barcodes and double-digested data. Programs now support a mixture of
	barcodes from single-end inline or index barcodes, to mixtures of inline/index barcodes.
	    1) changed naming scheme for process_radtags/process_shortreads output files for
	paired reads. Changed file suffix to properly be ".fq" or ".fa", with paired-reads named
	sample_XXX.1.fq and sample_XXX.2.fq instead of the previous ".fq_1" and ".fq_2".
	    2) Paired-reads remain synced in output files, with sinlgetons written to
	sample_XXX.rem.1.fq and sample_XXX.rem.2.fq.
	    2) changed Phred+33 to be the default encoding scheme (previously was the now
	deprecated Phred+64)
	    3) Combinatorial barcdoes are specified as --inline_index or --inline_inline among a
	number of other supported possibilities. Barcodes are listed in the barcode file as either
	a single column or two, tab-separated columns.
	    4) Two restriction enzymes can now be specified via --renz_1 and --renz_2 to have the
	program check (and correct) the restriction enzyme cut site on the first and second read
	respectively.
	    5) programs now properly ignore files starting with "." which is required for Mac
	OS X's ".DS_Store" files and for "." and ".." on Linux.
	Bugfix: processing paired-end data with process_radtags could incorrectly alter the first
	few nucleotides of the paired-read when correcting barcodes.
	Bugfix: two regressions were fixed in process_shortreads causing all reads to be
	improperly trimmed.
	Bugfix: VCF output did not include sites fixed within and variable among populations.
	Bugfix: changed the parsing code to accept a wider range of Illumina named, paired-end
	files in process_radtags/shortreads.
	Bugfix: gzipped files were not read properly in process_radtags/shortreads when a directory
	was specified with -P.
	Bugfix: setting secondary read distance to 0 in ustacks (-N) was ineffective.
	Bugfix: changed the PHP code to remove 'Strict Standards' warnings and a few other warnings.
	Thanks to Yue Yu for tracking down the proper changes to avoid the warnings.

Stacks 0.99995 - February 19, 2013
----------------------------------
	Feature: added support for using Google's Sparsehash Object: http://code.google.com/p/sparsehash/
	If enabled at compile time, this object will replace all the hash maps with Google's sparsehash
	saving significant memory.
	Feature: removed the -S command line option from cstacks and sstacks. These programs now read this ID
	directly from the Stacks input files.
	Feature: altered ustacks to no longer store FASTQ/FASTA IDs from input files in memory to lower
	memory usage. Instead, an integer representing the read is stored and the IDs are read back in from
	disk just before results are written.
	Feature: added the '--write_single_snp' option to populations. When writing Genepop or Structure files
	this option will cause populations to write just the first SNP per locus to the file, avoiding potential
	problems with linked SNPs originating from the same locus.
	Feature: compressed the Hval/Stack/Rem objects to remove convenience integer variables to save memory.
	Feature: updated Stacks programs to use the newer TR1 unordered_map hash object instead of the
	deprecated SGI hash_map object.
	Bugfix: fixed a memory leak in cstacks in which not all of the Locus Class elements was being properly
	freed (only the SNP objects were being freed).
	Bugfix: Added code to denovo_map.pl/ref_map.pl to remove from the logfile the 'counter' lines that
	printed when initially loading radtags data.

Stacks 0.99994 - February 12, 2013
----------------------------------
	Bugfix: process_radtags/process_shortreads, when adding support for reads of different length, I
	clobbered the sequence truncation option. Fixed this regression.
	Bugfix: the kernel smoothing algorithms for calculating Fst, Pi, and Fis could sometimes segfault
	as some RAD sites can overlap. Added code to find and describe overlapping RAD sites and report these
	to the user.

Stacks 0.99993 - January 30, 2013
---------------------------------
	Feature: process_radtags/process_shortreads/ustacks can now read gzipped Fasta/Fastq input files.
	Feature: ref_map.pl/pstacks now supports the use of BAM alignment files. This feature is optional and must
	be enabled during compilation. It requires the Samtools library to be installed.
	Bugfix: When using referenced aligned data, soft-masked alignments (Ns) were getting imporperly injected into
	the SNP models, which would call them as Homozygous Ns, and this data would eventually be passed to the
	summary statistics in populations, which would make errant Fst calculations.
	Bugfix: In rare cases, sequences aligned to the negative strand had their base pair positions slightly off,
	this could cause a segfault during populations' kernel-smoothed Fst calculations.
	Bugfix: In populations, fixed a rare, infinite loop condition in Fisher's exact test for Fst calculations.
	Could occur due to a floating point rounding error when calculating allele frequencies for Fst calculation.

Stacks 0.99992 - January 8, 2013
--------------------------------
	Bugfix: floating point command line options were not being processed correctly and may have been
	truncated.

Stacks 0.99991 - December 17, 2012
----------------------------------
	Feature: process_shortreads and process_radtags can now filter for adapter sequence in raw data, trimming
	(process_shortreads) or discarding (process_shortreads/process_radtags) it. Mismatches to the adapter
	sequence are allowed to accomodate for sequencing error.
	Bugfix: added --merge flag to process_shortreads/process_radtags to handle regression where unbarcoded
	data should be merged together into single output files.
	Bugfix: code in cstacks to characterize differentially fixed SNPs was only running with -n > 0, but
	should also run by default if -g is specified.
	Feature: made automated correction thresholds for the genotypes program accessible from the command
	line, including --min_hom_seqs, --min_het_seqs, and --max_het_seqs options.
	Feature: refactored clone_filter to be more functional. Now can output sequences in FASTA or FASTQ
	(FASTA will save memory). Keeps sequence headers intact, can capture discarded reads, and prints
	a distribution of the number of cloned read pairs.
	Bugfix: Remainder reads weren't being written properly as the file handles weren't properly closed.
	Bugfix: Processing paired reads with process_radtags/process_shortreads was not functioning correctly,
	barcode was not being transferred properly from P1 to P2 read. Regression introduce Aug 21, 2012.
	Feature: added support for OneMap CP map export in genotypes.
	Bugfix: Fixed some bugs in pstacks/ustacks command line processing involving --alpha and --model_type.
	Bugfix: several bugs in the exact and approximate bootstrap algorithms were corrected. These algorithms
	are now robust.
	Bugfix: Added code to ensure command line IDs are in fact integers.
	Bugfix: fixed nucleotide positions were not being tallied across populations properly resulting in an
	incorrect value for number of sites and percent polymorphic sites in the sumstats_summary file.
	Bugfix: pstacks could identify a locus that despite having SNPs would have no haplotypes generated.
	This would late cause sstacks to segfault. Added code in pstacks to blacklist these loci and code
	in sstacks to catch this case and not segfault, now will print a warning.

Stacks 0.9999 - October 03, 2012
--------------------------------
	Feature: two bootstrapping procedures have been introduced into the populations program to
	determine the statistical significance of kernel smoothed windows. These algorithms are controlled
	by the --bootstrap and --bootstrap_reps command line options.
	Feature: summary summary statistics are now written for all populations, giving the mean, variance,
	and standard error for each of the population-specific summary stats. In addition, private alleles
	are identified and marked in the sumstats file, and summarized across populations. Number and
	percent of polymorphic loci are also reported. The actual variable nucleotides at each site are now
	reported in the sumstats file.
	Feature: the populations program can now generate kernel-smoothed values for Fis and Pi, in addition
	to the current support for Fst.
	Feature: the populations program can now output SNP data for use in the program Structure.
	Feature: various sections of the populations program have been parallelized.
	Feature: the populations program can now output SNP data in the Phylip file format. If --phylip is
	specified, the populations program will identify SNPs that are fixed within populations, but variable
	between populations and output these in a Phylip file. This file can then be fed into any phylogenetics
	program, such as PhyML. This feature is equivalent to the analysis done in Emerson, et al., 2010. In
	addition, if the --phylip_var flag is specified as well, variable sites within populations are encoded
	into the Phylip file using standard alternative nucleotide encodings.
	Feature: for ustacks/pstacks, the alpha significance level can now be specified on the command line.
	Specifying --alpha to ustacks or pstacks will set the chi square significance level to determine
	whether a heterozygous or homozygous model call is statistically significant. Legal values of alpha are
	0.1, 0.05 (the previous default), 0.01, or 0.001.
	Feature: for ustacks/pstacks, a new bounded SNP calling model has been introduced, allowing limits to
	be set on the error rate. This model allows the calling of SNPs to be affected by prior knowledge
	as to how likely polymorphism is in the data set. This behavior is controlled by the --bound_low and
	--bound_high parameters to ustacks and pstacks.
	Feature: additional sections of ustacks has been parallelized. In addition, stack merging has been
	changed to occur in a single step (instead of in rounds as done previously).
	Feature: the deleveraging algorithm in ustacks has been replaced with a simple algorithm
	based on a minimum spanning tree. A new parameter has been introduced, --max_locus_stacks,
	which controls the number of stacks allowed to be merged together into a single locus. Loci that
	contain more than --max_locus_stacks stacks are set aside and not added to the catalog later on.
	Feature: export_sql.pl now has two depth parameters, allele and locus depth, allowing for the filtering
	of loci based on either one.
	Feature: added a 'dry run' flag (-d) to denovo_map.pl and ref_map.pl to allow the pipeline to be tested
	to see what it would execute, before actually executing any programs.
	Bugfix: problem with the FASTA parser fixed (it was introduced with fixes to handle windows-style files).
	Bugfix: sample counts where off in batch_*.haplotypes.tsv file generated by populations program.

Stacks 0.9996 - August 24, 2012
-------------------------------
	Bugfix: fixed significant memory leak in Kmer hashing for both ustacks and cstacks. Results in an
	approximately 3.4x reduction in memory use for cstacks, and an approximately 1.6x reduction in
	ustacks.
	Feature: process_radtags and process_shortreads can handle non-Illumina FASTQ headers (any generic FASTQ type).
	Feature: process_radtags can process data without barcodes.
	Feature: process_radtags and process_shortreads can handle Illumina barcodes, when the barcode is not
	inline but is instead provided in the FASTQ header.
	Bugfix: Corrected the behavior of the '-m' parameter to populations and genotypes. It is meant to apply
	to the total depth of a stack at a locus, but was instead being applied to the depth of each allele at
	each locus.
	Feature: process_radtags and process_shortreads can now automatically discard reads marked as
	'failed' by Illumina's chastity/purity filter.
	Feature: added ecoT22I, mluCI, nlaIII, and sphI restriction enzymes to process_radtags
	Bugfix: modified Stacks programs to handle Windows-style line endings ('\r\n') from FASTQ, FASTA, and
	SAM files as well as population maps.
	Bugfix: changed populations' genepop output to only include loci that are variable in the populations
	specified. Previously, in some cases, additional fixed loci were included, which are not included in the
	VCF output, causing the two files to have different loci present.
	Bugfix: expected homozygosity and observed homozygosity were not being reported correctly in the sumstats
	files. The other population statistics were not affected by the bug.
	Feature: process_radtags and process_shortreads now print command and time executed to log file.

Stacks 0.9995 - July 05, 2012
-----------------------------
	Bugfix: Fst summary matrix was being incorrectly written.

Stacks 0.9994 - July 01, 2012
-----------------------------
	Feature: the populations program can now write a file in the GenePop format. GenePop files can be read
	by the GenePop program and converted for other population genetics programs such as Arlequin. Caution: you
	may not be able to include all loci from a Stacks run in the output as these programs aren't necessarily
	capable of handling such a volume of data. However, you can use populations' whitelist feature to only
	include certain loci in the output.
	Feature: the populations program now writes an Fst summary file providing a matrix of mean Fst measures
	for each pair of populations in the analysis.
	Feature: added two filters to populations to require a locus to be present in a certain percentage of
	individuals in a population, and requiring a locus to be present in a certain number of populations. If
	the former criteria is not reached, the locus is zeroed out only in the specific population, if the latter
	criteria is not met, the locus is discarded from the analysis.
	Feature: three Fst corrections are now provided by the populations program: requiring a locus to have a significant
	p-value (smaller than 0.05, although its configurable), applying a Bonferroni correction according to the number
	of data points in the sliding window, and applying a Bonferroni correction according to the number of data points
	in the genome. Loci that fail to reach statitical significance in each case are considered not different from zero
	and are set to zero.
	Feature: a filter can be specified to the populations program requiring a minimum allele frequency (MAF) at
	a locus to consider the locus variable. If an allele at a locus is below the MAF, the locus is considered fixed.
	Feature: when using a reference genome, Stacks can now work with samples of different sequence lengths.
	This means one can combine samples generated from different Illumina runs of different length. Each
	individual sample must be of the same length internally, however.
	Feature: pstacks can now handle gapped alignments properly. It parses the CIGAR string in the SAM file
	and inserts/removes Ns to accomodate indels and soft-masked alignment fragments. This prevents the SNP
	model from mistakenly calling polymorphisms due to indel frameshifts.
	Bugfix: Removed O(n^2) algorithm from Sliding window Fst calculation in populations program, significant
	speedup acheived.
	Bugfix: Updated load_radtags.pl to support population types and to import sumstats, fst, and genotypes
	files.
	Bugfix: fixed a small memory leak in DNANSeq.

Stacks 0.9993 - June 07, 2012
--------------------------------
	Feature: Added Fisher's Exact Test statistics to Fst estimates. This provides a p-value, an odds ratio
	along with a 95% confidence interval and a Log of Odds (LOD) score for each Fst estimate. These
	statistics allow one to decide if a particular Fst measurement is significant.
	Feature: denovo_map.pl and ref_map.pl now import population statistics files into the database (fst
	and sumstats files).
	Feature: Web interface now displays summary statistics and Fst values for every locus.
	Feature: population names can now be directly added through the web interface and they will be stored
	in the database and propogated.

Stacks 0.9992 - May 22, 2012
--------------------------------
	Bugfix: fixed massive memory leak in Fst calcuations in populations program.
	Bugfix: if using a population map to calculate Fst in the populations program, some individuals could
	be inadvertently attributed to the wrong populations, due to a mismatch between the indices of the
	population map (PopMap.h) and the indexes recorded for making the population summary (PopSum.h).
	Feature: population map can now be specified to denovo_map.pl and ref_map.pl. This data is
	populated into the database and samples are displayed according to their population in the web interface.
	Feature: improved denovo_map.pl and ref_map.pl to check for existence of input files.
	Bugfix: export_sql.pl wasn't properly using the new filters that use a lower and upper bound (snps, alle,
	pare).
	Feature: improved how values are generated for web-based filters, allowing for larger populations/maps.
	Improved HTML rendering for extremely long haplotype strings.
	Bugfix: corrected alleles to be output as "unphased" in VCF file; corrected homozygotes to be printed as
	diploid values, e.g. '0/0' or '1/1' instead of just '0'.
	Bugfix: changed reporting of SNPs on samples.php page to specify total number of SNPs and the number
	of polymorphic loci (containing one or more SNPs).
	Bugfix: an extra tab was being placed in the VCF output file.
	Feature: added flag to process_radtags to disable checking the integrity of the RAD site in each raw
	read. Added a flag to allow more nucleotide mismatches in the barcode when rescuing barcodes.

Stacks 0.9991 - April 17, 2012
--------------------------------
	Bugfix: replaced bit-rotted code causing all nucleotides to be masked as 'N' when fixed model engaged
	on ustacks.

Stacks 0.999 - April 11, 2012
--------------------------------
	Feature: Added support for the 1000 Genomes Project, Variant Call Format (VCF) in the populations
	program. (http://www.1000genomes.org/node/101). This file output includes the genotype calls for
	every individual for each locus, allele depth, and likelihood values for heterozygous SNP calls.

	Feature: implemented a three-bit compression scheme so that uncalled bases ('N's) can be stored
	in compressed format in pstacks. Other stacks programs currently use two-bit compression which is
	more compact, but can only store plain nucleotides ('A', 'C', 'G', 'T'). This restores earlier behavior
	that allowed Ns in pstacks prior to the implementation of the two-bit compression scheme.

	Bugfix: the populations program was only outputing sites to the summary statistics file (*.sumstats.tsv)
	if they were heterozygous in a population. This could give the impression that the same site may be
	absent in other populations when in reality it was simply fixed in the other populations. Now, if a
	site is heterozygous in any of the populations, it will be output for all populations.
	Bugfix: added lots of error checking code to populations so it properly handles
	poorly formatted population maps, missing files, and similar errors.
	Bugfix: added uncalled bases ('n', 'N', and '.') to the reverse complement function (reads
	aligned on the negative strand and processed by pstacks will be stored reverse complement.
	Bugfix: updated the PHP code as well as export_sql.pl to properly use the new filters for
	chromosome, basepair, as well as lower and upper ranges to various filters.
	Other: Removed the deprecated markers.pl, genotypes.pl, and process_radtags.pl programs from the distribution.

Stacks 0.998 - January 06, 2012
--------------------------------
	Feature: Pipeline is now aware if samples are submitted as a 'population' or a 'mapping cross'.
	A new command line option, -s, has been added to denovo_map.pl and ref_map.pl that will label
	the dataset as a population. The -p/-r flags continue to keep the samples as a mapping cross.
	Feature: The web interface has been updated to display more information specific to populations.
	The filtering code has been changed to include lower and upper limits for filter fields such
	as SNPs, alleles, and number of parents/samples.

	Feature: A new program, populations, has been written to be executed in place of the exisiting
	genotypes program when a population is being processed through the pipeline. A map specifiying
	which individuals belong to which population is submitted to the program and the program will then
	calculate population genetics statistics, expected/observed heterzygosity, Pi, and Fis at each
	nucleotide position.
	Feature: the populations program will compare all populations pairwise to compute Fst. If a set
	of data is reference aligned, then a kernel-smoothed Fst will also be calculated.
	     These statistics were originally designed by Paul Hohenlohe and Bill Cresko, and are
	     described in the paper: Population Genomics of Parallel Adaptation in Threespine Stickleback
	     using Sequenced RAD Tags,
	     http://www.plosgenetics.org/article/info%3Adoi%2F10.1371%2Fjournal.pgen.1000862
	     They have been implemented independently in Stacks.

	Feature: added the DpnII enzyme to the process_radtags program.
	Feature: Added new 'model' line to *.tags.tsv files. This line records the output of the SNP
	model at every position in the read as either Homozygous (O), Heterozygous (E), or unknown (U).
	Previously only polymorphic loci were recorded in the SNPs file (and this remains unchanged). The
	model output line is now also available in the web interface.

	Bugfix: fixed crasher bug in cstacks when parallel processing was enabled for genomic-aligned data.
	Bugfix: allele depths are now properly reported in reference-aligned data.

Stacks 0.997 - November 22, 2011
--------------------------------
	Feature: new program, called clone_filter, that will take a set of paired-end reads and
	reduce them according to PCR clones (a PCR clone is a pair of reads that match exactly,
	while paried-end reads from two different DNA molecules will nearly always be slightly
	different lengths).
	Feature: new program, called kmer_filter, that allows paired or single-end reads to be
	filtered according to the number or rare or abundant kmers they contain. Useful for both
	RAD datasets as well as randomly sheared genomic or transcriptomic data.
	Feature: new program, called process_shortreads, performs the same task as process_radtags
	for fast cleaning of randomly sheared genomic or transcriptomic data (a 'beta' version of
	this program has actually been distributed in the last few Stacks releases).

	Feature: the Stacks tags.tsv file format has a new column to record the DNA strand that a
	particular read is aligned to, currently only used in datasets aligned to a reference genome.
	Feature: pstacks now reverse complements all stacks aligned to the negative strand and
	stores them in this orientation in the output files and database. (All aligners always present
	these reads in the positive orientation.) This change allows one to align reads to a reference
	genome using a gapped aligner, such as Tophat or GSNAP and have the RAD site still align with
	genomic data. (One can then compare genomic RAD tags along with cDNA RAD tags.)

	Feature: added the '-d' flag to export_sql.pl to export allele depths from the database.
	Feature: altered process_radtags to store orphaned, paired-end reads in a remainder file,
	keeping paired-reads in frame.
	Bugfix: fixed the handling of the paired-end barcode in process_shortreads, added a check
	to make sure the barcodes from both pairs of a read match.
	Bugfix: genotypes was not capitalizing auto-corrected genotypes in the generic format (it
	was in joinmap/rqtl specific formats).
	Bugfix: corrected cut site sequence for ApeKI in process_radtags.
	Bugfix: process_radtags inadvertantly used newly initialized memory that had not been
	cleared, causing rare parsing errors when uncleared memory resembled portions of a FASTQ record.
	Bugfix: the default MySQL permissions were not being properly passed to index_radatags.pl.
	Bugfix: changed load_radtags.pl to extract parental IDs from directly catalog files, instead of
	relying on file names.
	Feature: added a 'dry run' option to load_radtags.pl so it will print what it intends to do
	without actually doing it.

Stacks 0.996 - October 5, 2011
---------------------------------
	Web interface updates:
	  * If the RAD tags are aligned to a reference genome, a filter is now available to view markers
	    from a particular genomic region.
	  * The individual RAD tag viewer now scrolls while keeping the scale view and consensus sequence
	    always visible.
	  * The RAD tag viewer now highlights columns for which the catalog locus shows a SNP, but the
	    RAD tag does not.
	  * In the genotype viewer, the map between the haplotype and genotype is now available.
	  * The depth of each RAD tag is now visible in the genotype viewer.
	  * The genotype viewer has now been integrated with the observed haplotype viewer. You can
	    make changes/corrections to genotypes directly now, no need to submit a form and wait for
	    the page to reload.
	Bugfix: process_radtags wasn't properly parsing the names of v1 Illumnina BUSTARD files.
	Bugfix: process_radtags wasn't counting the total number of barceded paired-end reads correctly.
	Bugfix: sstacks' impute_haplotype() was causing spurious matching in some, error-based cases.
	Bugfix: build system was not properly replacing the _PKGDATADIR_ variable in denovo/ref_map.pl
	programs.

Stacks 0.995 - September 23, 2011
---------------------------------
	Feature: sstacks can now handle samples and catalogs that have different length reads.
	Each individual sample must be constructed from the same length reads (by ustacks and cstacks),
	but between samples there can be different lengths, e.g. a catalog of length 50bp and samples
	of length 100bp, or vice versa.
	Feature: Added the ApeKI restriction enzyme to process_radtags
	Feature: process_radtags can now capture discarded reads to a file.
	Bugfix: a coding limitation was removed that required polymorphic sites in the catalog to
	contain only two alleles. Now, all four alleles can be recorded at a single site in a locus in
	the catalog.
	Bugfix: Exporting results from the web interface was not including manual genotype corrections
	when requested.

Stacks 0.994 - August 08, 2011
------------------------------
	Feature: added catalog index structure to cstacks to speed construction of catalog
	when using reference-aligned sequences.
	Feature: added a new output type, 'genomic' to genotypes. Outputs SNPs individually,
	encoded as a set of integers, for reference-aligned reads.
	Bugfix: pstacks was not writing individual stack sequences properly.
	Bugfix: process_radtags was still checking the quality of sequence that was
	destined to be truncated off the read.
	Bugfix: process_radtags segfault fixed, parsing stop position
	mis-specified in parse_input_record().

Stacks 0.993 - August 05, 2011
------------------------------
	Memory usage optimization: Individual sequence reads are now stored internally
	using a 2-bit encoding of DNA nucleotides. Some simple benchmarking of
	ustacks (previous version / new version):

	 Sample size         Elapsed Time       Used Memory
	-------------     -----------------    -------------
	 3.78m reads         3:16 /    3:23     4.64G /  1.86G
	17.62m reads      1:31:21 / 1:43:54    55.55G / 45.42G

	Feature: Added the programs sort_read_pairs.pl, exec_velvet.pl, load_sequences.pl
	to facilitate the assembly of paired-end RAD-Tags into mini-contigs and allow them
	to be uploaded into and viewed from the web interface.
	Bugfix: made process_radtags emit an error when an unrecognized
	restriction enzyme is specified.
	Bugfix: made process_radtags accept barcodes with trailing whitespace,
	such as would be seen in a DOS text file or if errant tabs are
	specified.

Stacks 0.992 - July 04, 2011
----------------------------
	Feature: process_radtags can now handle Phred+33 or Phred+64 encodings, Phred+33 is
	the new default encoding in Illumina's CASAVA software (v1.8).
	Bugfix: Changed the sql input parser to handle variable length input
	lines. Necessary if loading tens of individuals into a catalog.
	Bugfix: Added command line options to ustacks to better control the use of secondary reads
	in the stack-building procedure.

Stacks 0.991 - June 06, 2011
----------------------------
	Bugfix: genotypes was failing to parse Stacks output files with
	unanticipated names.
	Bugfix: when using ref_map.pl, tags without SNPs were failing to match
	against the catalog.

Stacks 0.99 - May 20, 2011
--------------------------
	*A new C++ genotypes program has been added. This program works independently from the
	database allowing the pipeline to fully function without installing the database. The
	new program performs the tasks once completed by markers.pl and genotypes.pl.
	    - The pipeline has been modified to now automatically execute the genotypes program
	      as the last stage in an analysis. It will generate a file containing the observed
	      haplotypes and an additional file containing a map-agnostic set of genotype calls.
	    - If SQL interaction is enabled, the genotypes will be imported to the database and
	      serve as a base to export genotypes directly from the web interface for a particular
	      map and using the set of filters available online.
	    - If a population is being examinined, the observed haplotypes file can be imported into
	      Microsoft Excel or another tab-separated file viewer to immediately see the results.
	    - By replacing the Perl version of genotypes.pl we also no longer need to install or
	      worry about the caching mechanism for auto-correcting stacks, the C++ version can do
	      this by directly reading the Stacks output files.
	*markers.pl and genotypes.pl are now deprecated and will no longer be supported.
	*Feature: When exporting observed haplotypes, you can now specify a
	minimum stack depth to include a particular individual at a locus.
	*Feature: map-specific genotypes can now be exported directly from the
	database/web server.
	*Bugfix: genotypes.pl: make script ignore parental genotypes based on
	the sample type from the MySQL table, not based on the file name.
	*Bugfix: genotypes.pl: some loci were sneaking in despite being under
	the progeny limit.
	*Bugfix: made process_radtags Bustard file parser check number of fields to prevent
	attempting to parse FASTQ (and segfaulting). Thanks to
	Maureen.Liu -at- nottingham.ac.uk for reporting it.
	*Bugfix: in sstacks, when matching to the catalog using reads aligned
	to a reference genome (-g), sstacks did not verify that haplotypes
	matched exactly, causing some spurious matching, which later
	translated into dropped genotypes.
	*Bugfix: in markers.pl, the ratio observed alleles in the progeny was
	not being tallied correctly for ab/ac markers.

Stacks 0.984 - May 04, 2011
---------------------------
	*Bugfix: renamed constants.php to constants.php.dist to avoid
	overwriting an existing file on reinstallation.
	*Feature: process_radtags has been converted to a C++ program
	increasing its speed by approximately 25x. The parameters were
	modified to be a little more intuitive and parameters were added to
	control the size and score limit of the sliding window. The program
	can process a GAII lane in about 5 minutes, a HiSeq lane in about 12
	minutes, depending on the hardware used.

Stacks 0.983 - Apr 30, 2011
---------------------------
	*Bugfix: sstacks segfault when running parallelized. Improper
	insertion into map object when it should have only been checking for
	element presence/absence. Thanks to <yannick -dot- wurm -at- unil -dot- ch>
	for first reporting it.
	*Feature: added code to impute the genotype of a missing, second
	parent for some map types. This code adds up all the observed
	haplotypes in the progeny and then compares their frequencies against
	those that would be expected for the marker under Hardy-Weinberg
	equilibrium, choosing the marker type that best fits the
	Hardy-Weinberg expectation.

Stacks 0.982 - Mar 29, 2011
---------------------------
	*Bugfix: process_radtags.pl was not properly parsing FASTQ formated,
	 paired-end file names.
	*Bugfix: counts of matching parents/progeny were sometimes incorrect
	 due to a slightly promiscuous SQL query in index_radtags.pl.

Stacks 0.98 - Feb 25, 2011
---------------------------
        Note: if you have pre-existing databases, you must rebuild the catalog
	index (index_radtags.pl -D db -c) so that they are compatible with
	the new elements of the web interface.

	*Added option to pstacks to require a minimum depth of coverage for
	 a stack aligned to the refernce genome before reporting it.
	*Added double haploid (DH) and F2 export types to the genotypes.pl
	 script.
	*Added option to output any map in R/QTL output in genotypes.pl
	*Added feature to filter by number of available genotypes in progeny
	*Added command line option to ustacks to capture and output unused
         reads.
	*Added display of chromosome/base pair to web interface for stacks
	 aligned to a reference genome.
	*Bugfix: FASTA parser was missing records due to a bug introduced from
	 a FASTQ parser fix.
	*Bugfix: process_radtags.pl was not properly checking the integrity of
	 the RAD site after adding restriction enzymes with alternate
	 nucleotides.
	*Bugfix: when constructing the catalog, some tags being added to the
	 did not have their genomic location transferred over to a new catalog
	 tag.
	*Modified sstacks to include an option to match stacks against the
	 catalog based on the genomic location (assuming individuals were
	 processed with pstacks).
        *Bugfix: Lots of clean-ups and command line option fixes, thanks
	 to <yannick -dot- wurm -at- unil -dot- ch>.

Stacks 0.971 - Jan 30, 2011
---------------------------
	*Illumina software version 1.3 produces Phred scores that can begin
	 with a '@' character, throwing off the FASTQ parser. Added code to
	 clear the read buffer in between records to solve the problem. Thanks
	 to Aarti <aartiv -at- uchicago -dot- edu> for finding the bug.

Stacks 0.97
---------------------------
	*ustacks now detects when there are uncalled nucleotides in FASTA or
	 FASTQ input files, replaces those bases with 'A'.
	*process_radtags.pl now detects barcode length automatically. Removed
	 spurious error messages when no data is processed.

Stacks 0.96 - Jan 7, 2011
---------------------------
	*Fixed typo in README giving the wrong file path for the Apache
	 configuration file.
	*Fixed several hard-coded paths in PHP files that referred to our local
	 system.