Prediction of multi-drug resistance transporters using a novel sequence analysis method

MDRpred

Code to predict multi-drug resistance transporters.

MDRPred version 1.0

Prediction of multi-drug resistance transporters from protein sequence

Requires PyPro (https://code.google.com/p/protpy/wiki/propy) and

biopython (http://biopython.org/)

Usage: MDRPred.py -p [pattern filename] -i [input fasta filename] -o [output filename] -h

pattern filename: A tab-delimited file that includes a PyPro CTD designation string

(column 1; NA if none) and a regular expression (column 2)

fasta filename: Input sequences to be searched

output filename: Output file. Format will be tab-delimited file with (column order):

1. sequence identifier

2. match count

3-38. Match score from each input pattern

39. Match score for each sequence location. 0-9+ indicates the range of

0-100% of the models matching that location.

For the MDRpred patterns as described in the paper (http://f1000research.com/articles/4-60/v1)

sequences with 36 matches are the highest confidence (positive predictive value ~0.3,

that is, 1 out of 3 positive predictions was a true positive in our hands)

Limitations:

The method as currently implemented does not predict if a protein is a transporter in general.

That is, a positive prediction by MDRpred should be examined by other means to determine if

it's likely to be a transporter-type protein. Spurious results might be output when applied

to non-transporter proteins. This will be fixed in future releases.