rpsblast-summary-V4.pl

The aim of this script is to screen for a set of sequence patterns in all genomes present in the current directory. Screens are performed using rpsblast (tested with version 2.8.1+, included in Debian package ncbi-blast+) and patterns can be obtained from the CDD database at NCBI webpage. Nevertheless, if using the complete database the program will take too long to run for a data base with thousands of genomes. In such cases, it is recommended to construct a specific database that only includes the patterns that should be found in the genes of interest, as well as those expected to be absent from those genes. Full instructions for constructing your databases can be found at the CDD database webpage at NCBI. Briefly, you should run the makeprofiledb program. For example, supposing you have already downloaded the patterns of interest:

makeprofiledb -in patterns_of_interest.txt -out database_name

where “patterns_of_interest.txt” is the file containing the names of the patterns of interest and “database_name” is the name your database will have.

Program will look for the indicated pattern(s) in all the files of the directory that follow the indicated rules. Information such as the patterns file and gene selection rules must be indicated in a . Please look at RPSB-criteria-example.txt for an example of . Additionally, it is necessary to indicate the base for the name of the results directory and for the results files. must end with "/".
Each gene in the analyzed genomes will be selected if it has any of the required patterns in the inclusion criteria
Each gene in the analyzed genomes will be rejected if it has any of the patterns in the rejection criteria

Usage: perl rpsblast-summary-V4.pl