Samifier

Tools to enable a nexus between proteomic and genomic analysis. See https://github.com/IntersectAustralia/ap11_samifier/wiki for building, deployment and user guide. For scientific background, see http://intersectaustralia.github.com/ap11/ for details.

The code is licensed under the GNU GPL v3 license - see LICENSE.txt

The documentation (contained in the Github wiki and this README) is licensed under Creative Commons Attribution-Share Alike

Building

$ ant dist

This builds 4 command line toold and 2 helpers (undocumented). The following describes briefly each tool parameters. We encourage users to download the user guide and read it.

Samifier

Converts a search result from the Mascot protein search engine (or compatible) into SAM format, so it can be displayed in a genomics viewer.

$ java -jar dist/samifier.jar 
usage: samifier  -r <searchResultsFile> -c <chromosomeDir> -g <genomeFile>   
       -m <mappingFile> -o <outputFile> [-l <logFile>] [-b <bedFile>]
       [-s <Confidence Score thresold>]
 -r <searchResultsFile>           Mascot search results file in txt format           
 -c <chromosomeDir>               Directory containing the chromosome
                                  files in FASTA format for the given
                                  genome
 -m <mappingFile>                 File mapping protein identifier to
                                  ordered locus name                                      
 -g <genomeFile>                  Genome file in gff format
 -o <outputFile>                  Filename to write the SAM format file to     
 -l <logFile>                     Filename to write the log into
 -b <bedFile>                     Filename to write IGV regions of
                                  interest (BED) file to
 -s <Confidence Score thresold>   Minimum confidence score for peptides to
                                  be included

E.g.

$ java -jar samifier.jar -r results.txt -c saccharomyces_cerevisiae -g saccharomyces_cerevisiae_R64-1-1_20110208.gff -m accession.txt -o test.sam

Results analyser

Similar to samifier but instead of generating a SAM file, it generates a column with found peptides. This table that can be queried using SQL to extract a number of reports.

$ java -jar dist/results_analyser.jar 
usage: result_analyser -c <chromosomeDir> -g <genomeFile> -m <mappingFile>
       -o <outputFile> -r <searchResultsFile> [-rep <reportId>] [-replist
       <reportList>] [-sql <sqlQuery>]
 -c <chromosomeDir>       Directory containing the chromosome files in
                          FASTA format for the given genome
 -g <genomeFile>          Genome file in gff format
 -m <mappingFile>         File mapping protein identifier to ordered locus
                          name
 -o <outputFile>          Filename to write the SAM format file to
 -r <searchResultsFile>   Mascot search results file in txt format
 -rep <reportId>          Access a built in report query
 -replist <reportList>    A file containing all the pre-built SQL queries
 -sql <sqlQuery>          Filters the result through the use of a SQL
                          statement to the output file

Protein generator

Having as input a genome, it generates a FASTA file with "proteins" suitable to be used as a database in Mascot. It operates in two modes, using Glimmer predicted genes, or simply by splitting the genome into overlaping regions of given length.

Both Predicted Protein Generator (Glimmer gene prediction) and Virtual Protein Generator (six-frame translation) are implemented under the command line tool ‘protein_generator.jar’ as both tools shares similar input files. However, the Predicted Protein Generator can be accessed via command line parameter ‘-g ’ to identify the input Glimmer prediction file. The Virtual Protein Generator is accessed using the command line parameter ‘-i ’, which indicates the length of the overlapping virtual proteins.

$ java -jar dist/protein_generator.jar
usage: protein_generator -d <Database Name> -f <Genome File> [-g <Glimmer
       File>] [-i <Split Interval>] -o <Output File> [-p <GFF File>] [-q
       <Accession File>] [-t <Translation Table File>]
 -d <Database Name>            Database name
 -f <Genome File>              Genome file in FASTA format
 -g <Glimmer File>             Glimmer txt file. Can't be used with the -i
                               option.
 -i <Split Interval>           Size of the intervals (number of codons)
                               into which the genome will be split. Can't
                               be used with the -g option.
 -o <Output File>              Filename to write the FASTA format file to
 -p <GFF File>                 Filename to write the GFF file to
 -q <Accession File>           Filename to write the accession file to
 -t <Translation Table File>   File containing a mapping of codons to
                               amino acids, in the format used by NCBI.

Virtual protein merger

When using the database generated by the Protein Generator in interval mode the generated "proteins" are most likely wrong. However, Mascot still uses them and can report back found peptides in such sequences. The Virtual Protein Merger takes such search result against a "virtual protein" database and tries to rebuild the intervals by searching for stop and start codons in the sequence.

$ java -jar dist/virtual_protein_merger.jar
usage: virtual_protein_merger -c <chromosomeDir> -g <genomeFile> -o
       <outputFile> -r <searchResultsFile> [-t <Translation Table File>]
 -c <chromosomeDir>            Directory containing the chromosome files
                               in FASTA format for the given genome
 -g <genomeFile>               Genome file in gff format
 -o <outputFile>               Filename to write the gff file to
 -r <searchResultsFile>        Mascot search results file in txt format
 -t <Translation Table File>   File containing a mapping of codons to
                               amino acids, in the format used by NCBI.