Protein threading with residue-environment matching by artificial neural networks

Jiang, Nan; Wu, Wendy; Mitchell, Ian

doi:10.1145/967900.967943

Cited by 2 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The structural environment description in this research is on amino acid residue level and follows our previous work [11]. Each amino acid residue is described as main-chain and pseudo side-chain spheres.…”

Section: Description Of Structural Environmentsmentioning

confidence: 99%

See 1 more Smart Citation

Protein Fold Recognition by Mixed Environment-Specific Amino Acid Substitution Mapping Using Neural Networks

Jiang¹,

Wu²,

Mitchell³

11th International Conference on Parallel and Distributed Systems (ICPADS'05)

View full text Add to dashboard Cite

Protein fold recognition programs align a probe amino acid sequence onto a library of representative folds of known structure to identify structural similarities. Substitution matrix is the key to detect the aligned protein similarities. In this paper, a new mixed environment-specific substitution mapping (MESSM) is designed for fold recognition. It has two features: first, with amino acid residue level environmental description, structurally-derived substitution scores are generated using neural networks. The substitution probability of each pair of amino acids at any chosen structural environment can be instantly generated; second, the structurally-derived substitution score is linearly combined with sequence profile from traditional sequences substitution matrices to obtain a positive consensus for fold recognition. By fitting a single parameter in the combined substitution score, benchmark problems have been carried out to test the ability of the MESSM model. The results show that the new fold recognition model with mixed substitution mapping has a better performance than the one with either structure or sequence profile only. Moreover, it is comparable with those more computational intensive, energy potential based fold recognition model.

show abstract

Section: Description Of Structural Environmentsmentioning

confidence: 99%

“…Twenty units are represented the type of under consideration, which is encoded by orthogonal encoding scheme. Based on twenty types of amino acids, the left twenty input units are employed to encode the neighbour contacts of a[11].…”

mentioning

confidence: 99%