Biomedical subjects

Da-Fu Ding

Publications and source records attributed to Da-Fu Ding.

23 records · Page 2Linked to original sources

PubMed1998

Predicting Binding Free Energy for Protein Complexes.

A novel empirical free energy function and a computationally more effect algorithm are applied to 21 protein complexes to predict the free energies of binding. Compared with the other works what have done in literature, better agreement between the predicted and measured binding free energies is obtained. The predicted values, are typically within 1.0 kcal/mol of the measured, while keeping a high correlation coefficient of 96 %. Predicting the binding free energy of a typical protease-inhibitor complex takes about 2 min on SGI-IMPACT R10000 Workstation. What is more, the results reconfirm that the interactions between the hydrophilic atoms may play a more favorable role in binding than in folding.

Journal Article↗

PubMed1998

An Assembly Algorithm for DNA Sequence with Repeats.

A program for assembling DNA fragments using a new approach has been developed. In the program, a filter and a sequence alignment are used to identify the true overlap between the two fragments. Then the fragments in repeat copies are put into a repeat contig and the others are put into nonrepeat contigs. The repeat contig is resolved into two contigs and then all contigs are merged into one contig. At last the multiple alignment is obtained by the order of fragments in the contig, thereby forming a consensus sequence.The performance tests of the program for eight fragment libraries have demonstrated the feasibility of the program.

Journal Article↗

PubMed1997

Detecting Repetitive DNA in Eukaryotes.

A detection software of repetitive elements, especially for Alu is designed according to the theory of the correspondence analysis. The results obtained on the test set of 38 gene sequences showed that the true-negative is about 5.8% and false-positive is about 4.8%. Also, the detailed description of Alu elements in the sequences in which the Alu elements are spliced into coding-regions and the successful search for Alu in large-scale T-cell receptor loci sequence showed that the software is a better tool for the recognition of repeat elements.

Journal Article↗

PubMed1996

Large-scale Homologous Analysis of Genome Sequence.

We described a new method for the large-scale homologous analysis of genome sequences, which used hashing technique combined with sparse dynamic programming to get a sequence alignment. Three examples, the plant chloroplast genomes, the mammalian T-cell receptors C(alpha)/C(delta) gene loci and the mammalian gamma-crystallin gene clusters were analysed. The results showed that the method was more rapid to obtain accurate enough data and might be useful in genome analysis.

Journal Article↗

PubMed1996

A Konowledge-based Protein Modeling System and Its Evaluation III. A Sensitive Method for Evaluation Protein Model.

A new scoring table of compatibilities of residues with environments in a protein has been derived from environment-dependent amino acid substitution tables. The environments of the residues in the 3D structure are described by: (1) the secondary structure; (2) the solvent accessibility; (3) the hydrogen bonding. On the basis of the scoring table, an assessment method of protein models is proposed. Compared with the famous software Profile-3D, The method not only shows a more effective test of the accuracy of a 3D protein model, but also exhibits a higher sensitivity to the local misfolding. It is an important module in our protein modeling system PMODELING.

Journal Article↗