Learning from past treatments and their outcome improves prediction of in Vivo response to anti-HIV therapy

Hiroto Saigo, Andre Altmann, Jasmina Bogojeska, Fabian Mller, Sebastian Nowozin, Thomas Lengauer

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Infections with the human immunodeficiency virus type 1 (HIV-1) are treated with combinations of drugs. Unfortunately, HIV responds to the treatment by developing resistance mutations. Consequently, the genome of the viral target proteins is sequenced and inspected for resistance mutations as part of routine diagnostic procedures for ensuring an effective treatment. For predicting response to a combination therapy, currently available computer-based methods rely on the genotype of the virus and the composition of the regimen as input. However, no available tool takes full advantage of the knowledge about the order of and the response to previously prescribed regimens. The resulting high-dimensional feature space makes existing methods difficult to apply in a straightforward fashion. The machine learning system proposed in this work, sequence boosting, is tailored to exploiting such high-dimensional information, i.e. the extraction of longitudinal features, by utilizing the recent advancements in data mining and boosting. When applied to predicting the latest treatment outcome for 3,759 treatment-experienced patients from the EuResist integrated database, sequence boosting achieved superior performance compared to SVMs with RBF kernels. Moreover, sequence boosting allows an easy access to the discriminative treatment information. Analysis of feature importance values provided by our model confirmed known facts regarding HIV treatment. For instance, application of potent and recently licensed drugs was beneficial for patients, and, conversely, the patient group that was subject to NRTI mono-therapies in the past had poor treatment perspectives today. Furthermore, our model revealed novel biological insights. More precisely, the combination of previously used drugs with their in vivo response is more informative than the information of previously used drugs alone. Using this information improves the performance of systems for predicting therapy outcome.

Original languageEnglish
Article number6
JournalStatistical Applications in Genetics and Molecular Biology
Volume10
Issue number1
DOIs
Publication statusPublished - Feb 16 2011
Externally publishedYes

Fingerprint

Viruses
Therapy
Learning systems
Learning
HIV
Patient treatment
Boosting
Prediction
Drugs
Data mining
Genes
Proteins
Therapeutics
Virus
Chemical analysis
Mutation
High-dimensional
Pharmaceutical Preparations
Learning Systems
Feature Space

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Molecular Biology
  • Genetics
  • Computational Mathematics

Cite this

Learning from past treatments and their outcome improves prediction of in Vivo response to anti-HIV therapy. / Saigo, Hiroto; Altmann, Andre; Bogojeska, Jasmina; Mller, Fabian; Nowozin, Sebastian; Lengauer, Thomas.

In: Statistical Applications in Genetics and Molecular Biology, Vol. 10, No. 1, 6, 16.02.2011.

Research output: Contribution to journalArticle

Saigo, Hiroto ; Altmann, Andre ; Bogojeska, Jasmina ; Mller, Fabian ; Nowozin, Sebastian ; Lengauer, Thomas. / Learning from past treatments and their outcome improves prediction of in Vivo response to anti-HIV therapy. In: Statistical Applications in Genetics and Molecular Biology. 2011 ; Vol. 10, No. 1.
@article{27ec8e3690b24355829961ba454b3f1a,
title = "Learning from past treatments and their outcome improves prediction of in Vivo response to anti-HIV therapy",
abstract = "Infections with the human immunodeficiency virus type 1 (HIV-1) are treated with combinations of drugs. Unfortunately, HIV responds to the treatment by developing resistance mutations. Consequently, the genome of the viral target proteins is sequenced and inspected for resistance mutations as part of routine diagnostic procedures for ensuring an effective treatment. For predicting response to a combination therapy, currently available computer-based methods rely on the genotype of the virus and the composition of the regimen as input. However, no available tool takes full advantage of the knowledge about the order of and the response to previously prescribed regimens. The resulting high-dimensional feature space makes existing methods difficult to apply in a straightforward fashion. The machine learning system proposed in this work, sequence boosting, is tailored to exploiting such high-dimensional information, i.e. the extraction of longitudinal features, by utilizing the recent advancements in data mining and boosting. When applied to predicting the latest treatment outcome for 3,759 treatment-experienced patients from the EuResist integrated database, sequence boosting achieved superior performance compared to SVMs with RBF kernels. Moreover, sequence boosting allows an easy access to the discriminative treatment information. Analysis of feature importance values provided by our model confirmed known facts regarding HIV treatment. For instance, application of potent and recently licensed drugs was beneficial for patients, and, conversely, the patient group that was subject to NRTI mono-therapies in the past had poor treatment perspectives today. Furthermore, our model revealed novel biological insights. More precisely, the combination of previously used drugs with their in vivo response is more informative than the information of previously used drugs alone. Using this information improves the performance of systems for predicting therapy outcome.",
author = "Hiroto Saigo and Andre Altmann and Jasmina Bogojeska and Fabian Mller and Sebastian Nowozin and Thomas Lengauer",
year = "2011",
month = "2",
day = "16",
doi = "10.2202/1544-6115.1604",
language = "English",
volume = "10",
journal = "Statistical Applications in Genetics and Molecular Biology",
issn = "1544-6115",
publisher = "Berkeley Electronic Press",
number = "1",

}

TY - JOUR

T1 - Learning from past treatments and their outcome improves prediction of in Vivo response to anti-HIV therapy

AU - Saigo, Hiroto

AU - Altmann, Andre

AU - Bogojeska, Jasmina

AU - Mller, Fabian

AU - Nowozin, Sebastian

AU - Lengauer, Thomas

PY - 2011/2/16

Y1 - 2011/2/16

N2 - Infections with the human immunodeficiency virus type 1 (HIV-1) are treated with combinations of drugs. Unfortunately, HIV responds to the treatment by developing resistance mutations. Consequently, the genome of the viral target proteins is sequenced and inspected for resistance mutations as part of routine diagnostic procedures for ensuring an effective treatment. For predicting response to a combination therapy, currently available computer-based methods rely on the genotype of the virus and the composition of the regimen as input. However, no available tool takes full advantage of the knowledge about the order of and the response to previously prescribed regimens. The resulting high-dimensional feature space makes existing methods difficult to apply in a straightforward fashion. The machine learning system proposed in this work, sequence boosting, is tailored to exploiting such high-dimensional information, i.e. the extraction of longitudinal features, by utilizing the recent advancements in data mining and boosting. When applied to predicting the latest treatment outcome for 3,759 treatment-experienced patients from the EuResist integrated database, sequence boosting achieved superior performance compared to SVMs with RBF kernels. Moreover, sequence boosting allows an easy access to the discriminative treatment information. Analysis of feature importance values provided by our model confirmed known facts regarding HIV treatment. For instance, application of potent and recently licensed drugs was beneficial for patients, and, conversely, the patient group that was subject to NRTI mono-therapies in the past had poor treatment perspectives today. Furthermore, our model revealed novel biological insights. More precisely, the combination of previously used drugs with their in vivo response is more informative than the information of previously used drugs alone. Using this information improves the performance of systems for predicting therapy outcome.

AB - Infections with the human immunodeficiency virus type 1 (HIV-1) are treated with combinations of drugs. Unfortunately, HIV responds to the treatment by developing resistance mutations. Consequently, the genome of the viral target proteins is sequenced and inspected for resistance mutations as part of routine diagnostic procedures for ensuring an effective treatment. For predicting response to a combination therapy, currently available computer-based methods rely on the genotype of the virus and the composition of the regimen as input. However, no available tool takes full advantage of the knowledge about the order of and the response to previously prescribed regimens. The resulting high-dimensional feature space makes existing methods difficult to apply in a straightforward fashion. The machine learning system proposed in this work, sequence boosting, is tailored to exploiting such high-dimensional information, i.e. the extraction of longitudinal features, by utilizing the recent advancements in data mining and boosting. When applied to predicting the latest treatment outcome for 3,759 treatment-experienced patients from the EuResist integrated database, sequence boosting achieved superior performance compared to SVMs with RBF kernels. Moreover, sequence boosting allows an easy access to the discriminative treatment information. Analysis of feature importance values provided by our model confirmed known facts regarding HIV treatment. For instance, application of potent and recently licensed drugs was beneficial for patients, and, conversely, the patient group that was subject to NRTI mono-therapies in the past had poor treatment perspectives today. Furthermore, our model revealed novel biological insights. More precisely, the combination of previously used drugs with their in vivo response is more informative than the information of previously used drugs alone. Using this information improves the performance of systems for predicting therapy outcome.

UR - http://www.scopus.com/inward/record.url?scp=79951491821&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79951491821&partnerID=8YFLogxK

U2 - 10.2202/1544-6115.1604

DO - 10.2202/1544-6115.1604

M3 - Article

VL - 10

JO - Statistical Applications in Genetics and Molecular Biology

JF - Statistical Applications in Genetics and Molecular Biology

SN - 1544-6115

IS - 1

M1 - 6

ER -