Award details

Protein structure prediction - development and benchmarking of machine algorithms

Reference	BIO12056
Principal Investigator / Supervisor	Professor Stephen Muggleton
Co-Investigators / Co-Supervisors
Institution	University of York
Department	Computer Science
Funding type	Research
Value (£)	79,844
Status	Completed
Type	Research Grant
Start date	01/02/2000
End date	01/02/2002
Duration	24 months

Abstract

A collaboration to apply existing new machine learning (ML) algorithms (Prof. Muggleton, York) on two protein structure prediction problems (Dr Sternberg, London). The protein problems are to learn predictive rules governing (i) the 3D fold of proteins and (ii) the conformation of loops. Existing ML algorithms are inductive logic programming (ILP) implemented in PROGOL (Muggleton), decision trees (CART, public domain), hidden Markov models (public domain). The new algorithm (Analogical Prediction) that is to be developed will learn from sparse data, a problem identified during our current Bioinformatics grant on using ILP on protein folds. The learnt rules will be embedded into protein structure prediction software. (Joint with grant BIO12010).

Summary

unavailable

Committee	Not funded via Committee
Research Topics	X – not assigned to a current Research Topic
Research Priority	X – Research Priority information not available
Research Initiative	Bioinformatics (Phase 2) (BIO) [1998-2000]
Funding Scheme	X – not Funded via a specific Funding Scheme

Associated awards:

BIO12010 Protein structure prediction - development and benchmarking of machine learning algorithms

I accept the terms and conditions of use (opens in new window)

export PDF file

back to list new search