BioSig
Genomic Sequence Indexing/Signature Generation and Searching.
A partial revive and simplification of Dr. Timothy Chappell's TopSig, focusing on genomic documents.
Getting Started
Utilising the Makefile, running make will build an executable file biosig that can be used as follows.
Indexing (index)
Generate a signature file from the given sequence file/s.
./biosig index [OPTIONS] sequenceFile [sequenceFile2 [...]] -o outfile.bsig
Available OPTIONS:
| Option | Default | Description |
|---|---|---|
| -kmerlen | 5 | Kmer length to hash. |
| -sigwidth | 1024 | Signature size in bits. |
| -sigdensity | 0.047619048 | Signature density 0-1. |
| -match | (disabled) | Match and store sequence IDs with the given regular expression. Will prioritise first group () if present. |
| -q | (disbaled) | Quiet operation; Suppress output. |
Indexing will result in the output of both a signature file outfile.bsig, as well as a header file outfile.bsig.head, with the following formats:
Signature File .bsig (continuous binary)
This file contains the signatures that were generated from the sequences.
[signature A][signature B][...]
Header File .head (newline-delimited plaintext)
This file contains the IDs of the sequences that were indexed. The first line contains metadata regarding the indexing parameters.
[kmer length],[signature width],[signature density],[sequence count]
[sequence id A]
[sequence id B]
[...]
Searching (search)
Search/Compare the signatures in a given signature file with one or more query signature files.
./biosig search [OPTIONS] targetSig.bsig querySig.bsig [querySig2.bsig [...]] -o resultfile
Available OPTIONS:
| Option | Default | Description |
|---|---|---|
| -threshold | 0.0 | Filter results lower than the given similarity threshold 0-1. |
| -top | 0 (disabled) | Retain only the top k results. |
| -format | tsv | Result output format. (tsv | csv | trec) |
| -unique | (disabled) | Do not compare signatures with identical IDs. |
| -match | (disabled) | Match and store sequence IDs with the given regular expression. Will prioritise first group () if present. |
| -q | (disbaled) | Quiet operation; Suppress output. |
Results will be output as follows, depending on the format:
tsv (tab-separated values)
[query] [target] [similarity]
csv (comma-separated values)
[query],[target],[similarity]
trec (Text REtrieval Conference)
[query] Q0 [target] [rank] [similarity] biosig