Supervised learning and prediction of physical interactions between human and HIV proteins
Identifieur interne : 002624 ( Main/Merge ); précédent : 002623; suivant : 002625Supervised learning and prediction of physical interactions between human and HIV proteins
Auteurs : Matthew D. Dyer [États-Unis] ; T. M. Murali [États-Unis] ; Bruno W. Sobral [États-Unis]Source :
- Infection, genetics and evolution : (Print) [ 1567-1348 ] ; 2011.
Descripteurs français
- Pascal (Inist)
- Wicri :
- topic : Homme.
English descriptors
- KwdEn :
Abstract
Background: Infectious diseases result in millions of deaths each year. Physical interactions between pathogen and host proteins often form the basis of such infections. While a number of methods have been proposed for predicting protein-protein interactions (PPIs), they have primarily focused on intra-species protein-protein interactions. Methodology: We present an application of a supervised learning method for predicting physical interactions between host and pathogen proteins, using the human-HIV system. Using a Support Vector Machine with a linear kernel, we explore the use of a number of features including domain profiles, protein sequence k-mers, and properties of human proteins in a human PPI network. We achieve the best cross-validation performance when we use a combination of all three of these features. At a precision value of 70% we obtain recall values greater than 40%, depending on the ratio of positive examples to negative examples used during training. We use a classifier trained using these features to predict new PPIs between human and HIV proteins. We focus our discussion on those predicted interactions that involve human proteins known to be critical for HIV replication and propagation. Examples of predicted interactions with support in the literature include those necessary for viral attachment to the host membrane and subsequent invasion of the host cell. Significance: Unlike intra-species PPIs, host-pathogen PPIs have not yet been experimentally detected on a large scale, though they are likely to play important roles in pathogenesis and disease outcomes. Computational methods that can robustly and accurately predict host-pathogen PPIs hold the promise of guiding future experiments and gaining insights into potential mechanisms of pathogenesis.
Links toward previous steps (curation, corpus...)
- to stream PascalFrancis, to step Corpus: 000063
- to stream PascalFrancis, to step Curation: 000031
- to stream PascalFrancis, to step Checkpoint: 000063
Links to Exploration step
Pascal:11-0292014Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Supervised learning and prediction of physical interactions between human and HIV proteins</title><author><name sortKey="Dyer, Matthew D" sort="Dyer, Matthew D" uniqKey="Dyer M" first="Matthew D." last="Dyer">Matthew D. Dyer</name><affiliation wicri:level="2"><inist:fA14 i1="01"><s1>Virginia Bioinformatics Institute, Virginia Tech, 1 Washington St.</s1><s2>Blacksburg, VA 24061</s2><s3>USA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>États-Unis</country><placeName><region type="state">Virginie</region></placeName></affiliation></author><author><name sortKey="Murali, T M" sort="Murali, T M" uniqKey="Murali T" first="T. M." last="Murali">T. M. Murali</name><affiliation wicri:level="2"><inist:fA14 i1="02"><s1>Department of Computer Science, Virginia Tech, 114 McBryde Hall</s1><s2>Blacksburg, VA 24061</s2><s3>USA</s3><sZ>2 aut.</sZ></inist:fA14><country>États-Unis</country><placeName><region type="state">Virginie</region></placeName></affiliation></author><author><name sortKey="Sobral, Bruno W" sort="Sobral, Bruno W" uniqKey="Sobral B" first="Bruno W." last="Sobral">Bruno W. Sobral</name><affiliation wicri:level="2"><inist:fA14 i1="01"><s1>Virginia Bioinformatics Institute, Virginia Tech, 1 Washington St.</s1><s2>Blacksburg, VA 24061</s2><s3>USA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>États-Unis</country><placeName><region type="state">Virginie</region></placeName></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">INIST</idno><idno type="inist">11-0292014</idno><date when="2011">2011</date><idno type="stanalyst">PASCAL 11-0292014 INIST</idno><idno type="RBID">Pascal:11-0292014</idno><idno type="wicri:Area/PascalFrancis/Corpus">000063</idno><idno type="wicri:Area/PascalFrancis/Curation">000031</idno><idno type="wicri:Area/PascalFrancis/Checkpoint">000063</idno><idno type="wicri:explorRef" wicri:stream="PascalFrancis" wicri:step="Checkpoint">000063</idno><idno type="wicri:doubleKey">1567-1348:2011:Dyer M:supervised:learning:and</idno><idno type="wicri:Area/Main/Merge">002624</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Supervised learning and prediction of physical interactions between human and HIV proteins</title><author><name sortKey="Dyer, Matthew D" sort="Dyer, Matthew D" uniqKey="Dyer M" first="Matthew D." last="Dyer">Matthew D. Dyer</name><affiliation wicri:level="2"><inist:fA14 i1="01"><s1>Virginia Bioinformatics Institute, Virginia Tech, 1 Washington St.</s1><s2>Blacksburg, VA 24061</s2><s3>USA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>États-Unis</country><placeName><region type="state">Virginie</region></placeName></affiliation></author><author><name sortKey="Murali, T M" sort="Murali, T M" uniqKey="Murali T" first="T. M." last="Murali">T. M. Murali</name><affiliation wicri:level="2"><inist:fA14 i1="02"><s1>Department of Computer Science, Virginia Tech, 114 McBryde Hall</s1><s2>Blacksburg, VA 24061</s2><s3>USA</s3><sZ>2 aut.</sZ></inist:fA14><country>États-Unis</country><placeName><region type="state">Virginie</region></placeName></affiliation></author><author><name sortKey="Sobral, Bruno W" sort="Sobral, Bruno W" uniqKey="Sobral B" first="Bruno W." last="Sobral">Bruno W. Sobral</name><affiliation wicri:level="2"><inist:fA14 i1="01"><s1>Virginia Bioinformatics Institute, Virginia Tech, 1 Washington St.</s1><s2>Blacksburg, VA 24061</s2><s3>USA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>États-Unis</country><placeName><region type="state">Virginie</region></placeName></affiliation></author></analytic><series><title level="j" type="main">Infection, genetics and evolution : (Print)</title><idno type="ISSN">1567-1348</idno><imprint><date when="2011">2011</date></imprint></series></biblStruct></sourceDesc><seriesStmt><title level="j" type="main">Infection, genetics and evolution : (Print)</title><idno type="ISSN">1567-1348</idno></seriesStmt></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>AIDS</term><term>Human</term><term>Human immunodeficiency virus</term><term>Molecular epidemiology</term><term>Predictive factor</term><term>Protein</term></keywords><keywords scheme="Pascal" xml:lang="fr"><term>SIDA</term><term>Facteur prédictif</term><term>Protéine</term><term>Homme</term><term>Virus immunodéficience humaine</term><term>Epidémiologie moléculaire</term></keywords><keywords scheme="Wicri" type="topic" xml:lang="fr"><term>Homme</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Background: Infectious diseases result in millions of deaths each year. Physical interactions between pathogen and host proteins often form the basis of such infections. While a number of methods have been proposed for predicting protein-protein interactions (PPIs), they have primarily focused on intra-species protein-protein interactions. Methodology: We present an application of a supervised learning method for predicting physical interactions between host and pathogen proteins, using the human-HIV system. Using a Support Vector Machine with a linear kernel, we explore the use of a number of features including domain profiles, protein sequence k-mers, and properties of human proteins in a human PPI network. We achieve the best cross-validation performance when we use a combination of all three of these features. At a precision value of 70% we obtain recall values greater than 40%, depending on the ratio of positive examples to negative examples used during training. We use a classifier trained using these features to predict new PPIs between human and HIV proteins. We focus our discussion on those predicted interactions that involve human proteins known to be critical for HIV replication and propagation. Examples of predicted interactions with support in the literature include those necessary for viral attachment to the host membrane and subsequent invasion of the host cell. Significance: Unlike intra-species PPIs, host-pathogen PPIs have not yet been experimentally detected on a large scale, though they are likely to play important roles in pathogenesis and disease outcomes. Computational methods that can robustly and accurately predict host-pathogen PPIs hold the promise of guiding future experiments and gaining insights into potential mechanisms of pathogenesis.</div></front></TEI><affiliations><list><country><li>États-Unis</li></country><region><li>Virginie</li></region></list><tree><country name="États-Unis"><region name="Virginie"><name sortKey="Dyer, Matthew D" sort="Dyer, Matthew D" uniqKey="Dyer M" first="Matthew D." last="Dyer">Matthew D. Dyer</name></region><name sortKey="Murali, T M" sort="Murali, T M" uniqKey="Murali T" first="T. M." last="Murali">T. M. Murali</name><name sortKey="Sobral, Bruno W" sort="Sobral, Bruno W" uniqKey="Sobral B" first="Bruno W." last="Sobral">Bruno W. Sobral</name></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Sante/explor/MersV1/Data/Main/Merge
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 002624 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Merge/biblio.hfd -nk 002624 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Sante
|area= MersV1
|flux= Main
|étape= Merge
|type= RBID
|clé= Pascal:11-0292014
|texte= Supervised learning and prediction of physical interactions between human and HIV proteins
}}
| This area was generated with Dilib version V0.6.33. |  |