Self-consistency test reveals systematic bias in programs for prediction change of stability upon mutation. / Usmanova, Dinara R.; Bogatyreva, Natalya S.; Ariño Bernad, Joan et al.
In: Bioinformatics, Vol. 34, No. 21, 01.11.2018, p. 3653-3658.Research output: Contribution to journal › Article › peer-review
}
TY - JOUR
T1 - Self-consistency test reveals systematic bias in programs for prediction change of stability upon mutation
AU - Usmanova, Dinara R.
AU - Bogatyreva, Natalya S.
AU - Ariño Bernad, Joan
AU - Eremina, Aleksandra A.
AU - Gorshkova, Anastasiya A.
AU - Kanevskiy, German M.
AU - Lonishin, Lyubov R.
AU - Meister, Alexander V.
AU - Yakupova, Alisa G.
AU - Kondrashov, Fyodor A.
AU - Ivankov, Dmitry N.
N1 - Publisher Copyright: © The Author(s) 2018. Published by Oxford University Press.
PY - 2018/11/1
Y1 - 2018/11/1
N2 - Motivation: Computational prediction of the effect of mutations on protein stability is used by researchers in many fields. The utility of the prediction methods is affected by their accuracy and bias. Bias, a systematic shift of the predicted change of stability, has been noted as an issue for several methods, but has not been investigated systematically. Presence of the bias may lead to misleading results especially when exploring the effects of combination of different mutations. Results: Here we use a protocol to measure the bias as a function of the number of introduced mutations. It is based on a self-consistency test of the reciprocity the effect of a mutation. An advantage of the used approach is that it relies solely on crystal structures without experimentally measured stability values. We applied the protocol to four popular algorithms predicting change of protein stability upon mutation, FoldX, Eris, Rosetta and I-Mutant, and found an inherent bias. For one program, FoldX, we manage to substantially reduce the bias using additional relaxation by Modeller. Authors using algorithms for predicting effects of mutations should be aware of the bias described here.
AB - Motivation: Computational prediction of the effect of mutations on protein stability is used by researchers in many fields. The utility of the prediction methods is affected by their accuracy and bias. Bias, a systematic shift of the predicted change of stability, has been noted as an issue for several methods, but has not been investigated systematically. Presence of the bias may lead to misleading results especially when exploring the effects of combination of different mutations. Results: Here we use a protocol to measure the bias as a function of the number of introduced mutations. It is based on a self-consistency test of the reciprocity the effect of a mutation. An advantage of the used approach is that it relies solely on crystal structures without experimentally measured stability values. We applied the protocol to four popular algorithms predicting change of protein stability upon mutation, FoldX, Eris, Rosetta and I-Mutant, and found an inherent bias. For one program, FoldX, we manage to substantially reduce the bias using additional relaxation by Modeller. Authors using algorithms for predicting effects of mutations should be aware of the bias described here.
KW - MODIFIED FOLDX PROTOCOL
KW - PROTEIN STABILITY
KW - SEQUENCE
KW - SERVER
UR - http://www.scopus.com/inward/record.url?scp=85053142079&partnerID=8YFLogxK
U2 - 10.1093/bioinformatics/bty340
DO - 10.1093/bioinformatics/bty340
M3 - Article
C2 - 29722803
AN - SCOPUS:85053142079
VL - 34
SP - 3653
EP - 3658
JO - Bioinformatics
JF - Bioinformatics
SN - 1367-4803
IS - 21
ER -
ID: 17248452