GeNT is a FORTRAN program that computes the cross entropy for groups of sequences that have been assigned to groups on the basis of biochemical,physiological, or other biological property. The sequence assignments are cross-validated, again by the cross entropy measure, to check for problems with the alignment or group
assignment. Sequences that were initially identified as “unclassified” are compared to all of the groups using position specific log-odds scores as described by Henikoff and Henikoff. Positions in the aligned sequences that are important for determining group membership are identified by having a high
entropy for the entire alignment and a high entropy for one or more specific groups.
Gent was originally written by Hugh Nicholas at the Pittsburgh Supercomputing Center.
Usage on Bridges-2
To see what versions of GeNT are available and if there is more than one, which is the default, along with some help, type
module spider gent
To use GeNT, include a command like this in your batch script or interactive session to load the GeNT module: (note ‘module load’ is case-sensitive):
module load gent
Sample Group file
Below is a sample group file. Sequences are from uniprot database:
Problem pa2_small Title PA2 Sequences. Group One sp|P00593|PA21B_BOVIN sp|P04417|PA21B_AGKHA sp|P14418|PA21B_AGKHP sp|P81243|PA21B_BOTJA sp|P0CAS3|PA212_CRODR sp|P0CAS4|PA213_CRODR sp|P0CAS5|PA215_CRODU sp|P0CAS6|PA216_CRODU sp|P0CAS7|PA217_CRODU // Group Two sp|P00606|PA20_BUNMU sp|Q8UW08|PA20_LAPHA sp|P08873|PA20_NOTSC sp|P20254|PA20_PSEAU sp|Q9PUH4|PA210_AUSSU sp|Q9PUH3|PA211_AUSSU sp|Q9PUH2|PA212_AUSSU sp|Q9PUH1|PA213_AUSSU sp|Q9PUH0|PA214_AUSSU sp|Q9PUG9|PA215_AUSSU sp|Q9PUG8|PA216_AUSSU sp|Q9PUG7|PA217_AUSSU sp|P81236|PA21B_ACAAN sp|P59359|PA21B_AUSSU sp|Q8QFW4|PA21B_BUNCE sp|Q90WA7|PA21B_BUNFA sp|P00617|PA21B_BUNMU //