Some regulatory datasets have a "score" column, generally indicating strength of the measured signal within genomic regions. Sometimes we want to consider only genomic regions with high strength signal, ignoring those with low signal.
The "score" generally ranges from 0 to 1000; however, there are multiple exceptions from this rule. Therefore, we implemented the percent (%) threshold filtering. For example, if for a given genome annotation track the score ranges from 0 to 1000, selecting 25% threshold will reomve genomic features with the score smaller than 250.
Filtering by score will creates the following folder structure:
/home/testuser/Documents/db_1.01_05.12.2014/grsnp_db_25
/home/testuser/Documents/db_1.01_05.12.2014/grsnp_db_50
/home/testuser/Documents/db_1.01_05.12.2014/grsnp_db_75
Filtering by strand further separates regulatory datasets into strand-specific datasets. This step creates the following folder structure:
/home/testuser/Documents/db_1.01_05.12.2014/grsnp_db_25_plus
/home/testuser/Documents/db_1.01_05.12.2014/grsnp_db_25_minus
Both filtering steps are invoked automatically by both dbcreator* modules.