An inductive logic programming approach to validate hexose binding biochemical knowledge

Houssam Nassif, Hassan Al-Ali, Sawsan Khuri, Walid Keirouz, David Page

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

Hexoses are simple sugars that play a key role in many cellular pathways, and in the regulation of development and disease mechanisms. Current protein-sugar computational models are based, at least partially, on prior biochemical findings and knowledge. They incorporate different parts of these findings in predictive black-box models. We investigate the empirical support for biochemical findings by comparing Inductive Logic Programming (ILP) induced rules to actual biochemical results. We mine the Protein Data Bank for a representative data set of hexose binding sites, non-hexose binding sites and surface grooves. We build an ILP model of hexose-binding sites and evaluate our results against several baseline machine learning classifiers. Our method achieves an accuracy similar to that of other black-box classifiers while providing insight into the discriminating process. In addition, it confirms wet-lab findings and reveals a previously unreported Trp-Glu amino acids dependency.

Original languageEnglish (US)
Title of host publicationInductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers
Pages149-165
Number of pages17
DOIs
StatePublished - Aug 3 2010
Externally publishedYes
Event19th International Conference on Inductive Logic Programming, ILP 2009 - Leuven, Belgium
Duration: Jul 2 2009Jul 4 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5989 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other19th International Conference on Inductive Logic Programming, ILP 2009
CountryBelgium
CityLeuven
Period7/2/097/4/09

Fingerprint

Inductive logic programming (ILP)
Inductive Logic Programming
Binding sites
Sugars
Black Box
Classifiers
Classifier
Proteins
Protein
Computational Model
Programming Model
Learning systems
Amino Acids
Amino acids
Baseline
Pathway
Machine Learning
Evaluate
Knowledge
Model

Keywords

  • Aleph
  • binding site
  • hexose
  • ILP
  • protein-carbohydrate interaction
  • rule generation
  • substrate recognition

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Nassif, H., Al-Ali, H., Khuri, S., Keirouz, W., & Page, D. (2010). An inductive logic programming approach to validate hexose binding biochemical knowledge. In Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers (pp. 149-165). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5989 LNAI). https://doi.org/10.1007/978-3-642-13840-9_14

An inductive logic programming approach to validate hexose binding biochemical knowledge. / Nassif, Houssam; Al-Ali, Hassan; Khuri, Sawsan; Keirouz, Walid; Page, David.

Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers. 2010. p. 149-165 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5989 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nassif, H, Al-Ali, H, Khuri, S, Keirouz, W & Page, D 2010, An inductive logic programming approach to validate hexose binding biochemical knowledge. in Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5989 LNAI, pp. 149-165, 19th International Conference on Inductive Logic Programming, ILP 2009, Leuven, Belgium, 7/2/09. https://doi.org/10.1007/978-3-642-13840-9_14
Nassif H, Al-Ali H, Khuri S, Keirouz W, Page D. An inductive logic programming approach to validate hexose binding biochemical knowledge. In Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers. 2010. p. 149-165. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-13840-9_14
Nassif, Houssam ; Al-Ali, Hassan ; Khuri, Sawsan ; Keirouz, Walid ; Page, David. / An inductive logic programming approach to validate hexose binding biochemical knowledge. Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers. 2010. pp. 149-165 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{70e1d2130740406bbc0a00a4a341b734,
title = "An inductive logic programming approach to validate hexose binding biochemical knowledge",
abstract = "Hexoses are simple sugars that play a key role in many cellular pathways, and in the regulation of development and disease mechanisms. Current protein-sugar computational models are based, at least partially, on prior biochemical findings and knowledge. They incorporate different parts of these findings in predictive black-box models. We investigate the empirical support for biochemical findings by comparing Inductive Logic Programming (ILP) induced rules to actual biochemical results. We mine the Protein Data Bank for a representative data set of hexose binding sites, non-hexose binding sites and surface grooves. We build an ILP model of hexose-binding sites and evaluate our results against several baseline machine learning classifiers. Our method achieves an accuracy similar to that of other black-box classifiers while providing insight into the discriminating process. In addition, it confirms wet-lab findings and reveals a previously unreported Trp-Glu amino acids dependency.",
keywords = "Aleph, binding site, hexose, ILP, protein-carbohydrate interaction, rule generation, substrate recognition",
author = "Houssam Nassif and Hassan Al-Ali and Sawsan Khuri and Walid Keirouz and David Page",
year = "2010",
month = "8",
day = "3",
doi = "10.1007/978-3-642-13840-9_14",
language = "English (US)",
isbn = "364213839X",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "149--165",
booktitle = "Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers",

}

TY - GEN

T1 - An inductive logic programming approach to validate hexose binding biochemical knowledge

AU - Nassif, Houssam

AU - Al-Ali, Hassan

AU - Khuri, Sawsan

AU - Keirouz, Walid

AU - Page, David

PY - 2010/8/3

Y1 - 2010/8/3

N2 - Hexoses are simple sugars that play a key role in many cellular pathways, and in the regulation of development and disease mechanisms. Current protein-sugar computational models are based, at least partially, on prior biochemical findings and knowledge. They incorporate different parts of these findings in predictive black-box models. We investigate the empirical support for biochemical findings by comparing Inductive Logic Programming (ILP) induced rules to actual biochemical results. We mine the Protein Data Bank for a representative data set of hexose binding sites, non-hexose binding sites and surface grooves. We build an ILP model of hexose-binding sites and evaluate our results against several baseline machine learning classifiers. Our method achieves an accuracy similar to that of other black-box classifiers while providing insight into the discriminating process. In addition, it confirms wet-lab findings and reveals a previously unreported Trp-Glu amino acids dependency.

AB - Hexoses are simple sugars that play a key role in many cellular pathways, and in the regulation of development and disease mechanisms. Current protein-sugar computational models are based, at least partially, on prior biochemical findings and knowledge. They incorporate different parts of these findings in predictive black-box models. We investigate the empirical support for biochemical findings by comparing Inductive Logic Programming (ILP) induced rules to actual biochemical results. We mine the Protein Data Bank for a representative data set of hexose binding sites, non-hexose binding sites and surface grooves. We build an ILP model of hexose-binding sites and evaluate our results against several baseline machine learning classifiers. Our method achieves an accuracy similar to that of other black-box classifiers while providing insight into the discriminating process. In addition, it confirms wet-lab findings and reveals a previously unreported Trp-Glu amino acids dependency.

KW - Aleph

KW - binding site

KW - hexose

KW - ILP

KW - protein-carbohydrate interaction

KW - rule generation

KW - substrate recognition

UR - http://www.scopus.com/inward/record.url?scp=77955027074&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77955027074&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-13840-9_14

DO - 10.1007/978-3-642-13840-9_14

M3 - Conference contribution

SN - 364213839X

SN - 9783642138393

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 149

EP - 165

BT - Inductive Logic Programming - 19th International Conference, ILP 2009, Revised Papers

ER -