=head1 NAME

ncbi-seg - segment sequence(s) by local complexity

=head1 SYNOPSIS

ncbi-seg sequence [ W ] [ K(1) ] [ K(2) ] [ -x ] [ options ]

=head1 DESCRIPTION

ncbi-seg divides sequences into contrasting segments of low-complexity
and high-complexity.  Low-complexity segments defined by the
algorithm represent "simple sequences" or "compositionally-biased
regions".

Locally-optimized low-complexity segments are produced at defined
levels of stringency, based on formal definitions of local
compositional complexity (Wootton & Federhen, 1993).  The segment
lengths and the number of segments per sequence are determined
automatically by the algorithm.

The input is a FASTA-formatted sequence file, or a database file
containing many FASTA-formatted  sequences.  ncbi-seg is tuned for amino
acid sequences.  For nucleotide sequences, see EXAMPLES OF
PARAMETER SETS below.

The stringency of the search for low-complexity segments is
determined by three user-defined parameters, trigger window length
[ W ], trigger complexity [ K(1) ] and extension complexity [ K(2)]
(see below under PARAMETERS ).  The defaults provided are suitable
for low-complexity masking of database search query sequences [ -x
option required, see below].


=head1 OUTPUTS AND APPLICATIONS

(1) Readable segmented sequence [Default].  Regions of contrasting
complexity are displayed in "tree format".  See EXAMPLES.

(2) Low-complexity masking (see Altschul et al, 1994).  Produce a
masked FASTA-formatted file, ready for  input as a query sequence for
database search programs such as BLAST or FASTA.  The amino acids in
low-complexity regions are replaced with "x" characters [-x option].
See EXAMPLES.

(3) Database construction.  Produce FASTA-formatted files containing
low-complexity segments [-l  option], or high-complexity segments
[-h option], or both [-a option].  Each segment is a separate
sequence entry with an informative header line.

=head1 ALGORITHM

The SEG algorithm has two stages.  First, identification of
approximate raw segments of low- complexity; second local
optimization.

At the first stage, the stringency and resolution of the search for
low-complexity segments is determined  by the W, K(1) and K(2)
parameters.  All trigger windows are defined, including overlapping
windows, of length W and complexity less than or equal to K(1).
"Complexity" here is defined by equation  (3) of Wootton & Federhen
(1993).  Each trigger window is then extended into a contig in both
directions by merging with extension windows, which are overlapping
windows of length W and complexity  less than or equal to K(2).
Each contig is a raw segment.

At the second stage, each raw segment is reduced to a single
optimal low-complexity segment, which  may be the entire raw
segment but is usually a subsequence.  The optimal subsequence has
the lowest  value of the probability P(0) (equation (5) of Wootton
& Federhen, 1993).

=head1 PARAMETERS

These three numeric parameters are in obligatory order after the
sequence file name.

Trigger window length [ W ].  An integer greater than zero [ Default
12 ].

Trigger complexity. [ K1 ].  The maximum complexity of a trigger
window in units of bits. K1 must  be equal to or greater than zero.
The maximum value is 4.322 (log[base 2]20) for amino acid
sequences [ Default 2.2 ].

Extension complexity [ K2 ].  The maximum complexity of an extension
window in units of bits.  Only values greater than K1 are effective
in extending triggered windows.  Range of possible values is as for
K1 [ Default 2.5 ].


=head1 OPTIONS

The following options may be placed in any order in the command
line after the W, K1 and K2 parameters:

=over

=item  -a

Output both low-complexity and high-complexity segments in a
FASTA-formatted file, as a set of  separate entries with header
lines.

=item  -c  [characters-per-line]

Number of sequence characters per line of
output [Default 60].  Other characters, such as residue numbers, are additional.

=item  -h

Output only the high-complexity segments in a FASTA-formatted
file, as a set of separate entries  with header lines.

=item  -l

Output only the low-complexity segments in a FASTA-formatted
file, as a set of separate entries with  header lines.

=item -m  [length]

Minimum length in residues for a high-complexity
segment [default 0].  Shorter segments are merged with adjacent
low-complexity segments.

=item -o

Show all overlapping, independently-triggered low-complexity
segments [these are merged by default].

=item  -q

Produce an output format with the sequence in a numbered block
with markings to assist residue counting.  The low-complexity and
high-complexity segments are in lower- and upper-case characters
respectively.

=item  -t  [length]

"Maximum trim length" parameter [default 100]. This
controls the search space (and  search time) during the
optimization of raw segments (see ALGORITHM above).  By default,
subsequences 100 or more residues shorter than the raw segment are
omitted from the search. This parameter may be increased to give
a more extensive search if raw segments are longer than 100 residues.

=item -x

The masking option for amino acid sequences.  Each input
sequence is represented by a single output sequence in FASTA-format
with low-complexity regions replaced by strings of "x" characters.

=back

=head1 EXAMPLES OF PARAMETER SETS

Default parameters are given by 'ncbi-seg sequence' (equivalent to 'ncbi-seg
sequence 12 2.2 2.5').  These  parameters are appropriate for low-
complexity masking of many amino acid sequences [with -x option  ].

=head2 Database-database comparisons:

More stringent (lower) complexity parameters are suitable when
masked sequences are compared with masked sequences.  For example,
for BLAST or FASTA searches that compare two amino acid sequence
databases, the following masking may be applied to both databases:

  ncbi-seg database 12 1.8 2.0 -x

=head2 Homopolymer analysis:

To examine all homopolymeric subsequences of length (for example)
7 or greater:

  ncbi-seg sequence 7 0 0

=head2 Non-globular regions of protein sequences:

Many long non-globular domains may be diagnosed at longer window
lengths, typically:

  ncbi-seg sequence 45 3.4 3.75

For some shorter non-globular domains, the following set is
appropriate:

  ncbi-seg sequence 25 3.0 3.3

=head2 Nucleotide sequences:

The maximum value of the complexity parameters is 2 (log[base 2]4).
For masking, the following is approximately equivalent in effect
to the default parameters for amino acid sequences:

  ncbi-seg sequence.na 21 1.4 1.6

=head1 EXAMPLES

The following is a file named 'prion' in FASTA format:

 >PRIO_HUMAN MAJOR PRION PROTEIN PRECURSOR
 MANLGCWMLVLFVATWSDLGLCKKRPKPGGWNTGGSRYPGQGSPGGNRYPPQGGGGWGQP
 HGGGWGQPHGGGWGQPHGGGWGQPHGGGWGQGGGTHSQWNKPSKPKTNMKHMAGAAAAGA
 VVGGLGGYMLGSAMSRPIIHFGSDYEDRYYRENMHRYPNQVYYRPMDEYSNQNNFVHDCV
 NITIKQHTVTTTTKGENFTETDVKMMERVVEQMCITQYERESQAYYQRGSSMVLFSSPPV
 ILLISFLIFLIVG

The command line:

 ncbi-seg __docdir__/examples/prion.fa

gives the standard output below


 >PRIO_HUMAN MAJOR PRION PROTEIN PRECURSOR

                                   1-49   MANLGCWMLVLFVATWSDLGLCKKRPKPGG
                                          WNTGGSRYPGQGSPGGNRY
 ppqggggwgqphgggwgqphgggwgqphgg   50-94
                gwgqphgggwgqggg
                                  95-112  THSQWNKPSKPKTNMKHM
        agaaaagavvgglggymlgsams  113-135
                                 136-187  RPIIHFGSDYEDRYYRENMHRYPNQVYYRP
                                          MDEYSNQNNFVHDCVNITIKQH
                 tvttttkgenftet  188-201
                                 202-236  DVKMMERVVEQMCITQYERESQAYYQRGSS
                                          MVLFS
               sppvillisflifliv  237-252
                                 253-253  G

The low-complexity sequences are on the left (lower case) and
high-complexity sequences are on the right (upper case).  All
sequence segments read from left to right and their order in the
sequence is from top to bottom, as shown by the central column of
residue numbers.

The command line:

  ncbi-seg __docdir__/examples/prion.fa -x

gives the following FASTA-formatted file:-

 >PRIO_HUMAN MAJOR PRION PROTEIN PRECURSOR
 MANLGCWMLVLFVATWSDLGLCKKRPKPGGWNTGGSRYPGQGSPGGNRYxxxxxxxxxxx
 xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxTHSQWNKPSKPKTNMKHMxxxxxxxx
 xxxxxxxxxxxxxxxRPIIHFGSDYEDRYYRENMHRYPNQVYYRPMDEYSNQNNFVHDCV
 NITIKQHxxxxxxxxxxxxxxDVKMMERVVEQMCITQYERESQAYYQRGSSMVLFSxxxx
 xxxxxxxxxxxxG

=head1 SEE ALSO

segn(1), blast(1), saps(1), xnu(1)

=head1 AUTHORS

John Wootton:     wootton@ncbi.nlm.nih.gov

Scott Federhen:   federhen@ncbi.nlm.nih.gov

 National Center for Biotechnology Information
 Building 38A, Room 8N805
 National Library of Medicine
 National Institutes of Health
 Bethesda, Maryland, MD 20894
 U.S.A.


=head1 PRIMARY REFERENCE

Wootton, J.C., Federhen, S. (1993)  Statistics of local complexity
in amino acid sequences and sequence  databases.  Computers &
Chemistry 17: 149-163.


=head1 OTHER REFERENCES

Wootton, J.C. (1994)  Non-globular domains in protein sequences:
automated segmentation using complexity measures.  Computers &
Chemistry 18: (in press).

Altschul, S.F., Boguski, M., Gish, W., Wootton, J.C. (1994)  Issues
in searching molecular sequence  databases.  Nature Genetics 6:
119-129.

Wootton, J.C. (1994)  Simple sequences of protein and DNA. In:
Nucleic Acid and Protein Sequence  Analysis: A Practical Approach.
(Second Edition, Chapter 8, Bishop, M.J. and Rawlings, C.R. Eds.
IRL  Press, Oxford) (In press).