Project titles for:
Project titles for: 2006-08
- Detection of putative miRNA sequences in rice genome
- QSAR Studies on Tuberculosis Inhibitors
- Eigenvalue analysis and its applications in Bioinformatics
- Parallel Implementation of DOCK Algorithm
- Comparative Analysis of Protein Classification Methods
- Identification of Neuraminidase inhibitors of Influenza virus
- Identification of Transcription factor binding sites in yeast
- Restriction site analysis of Rice genome
- Docking Studies on Tuberculosis Inhibitors
- Determination of sequence homology in promoter region for abiotic stress responsive genes in rice
- SNP mining in Chromosome no.8 of rice
- Annotation of EST Sequences present on Chromosome no. 1, 4 & 8 of Rice
- Function Prediction for genes near BAD-2 locus on Chromosome no. 8 of Rice
- Computational Prediction of Putative miRNA Candidates in Malarial Parasite Genome
- Multimedia database of fungal diseases in Rice and Wheat
- GePre: A Gene Prediction Tool
- SEALI: A Sequence Alignment Tool
- Tannase Structure Prediction
- Algorithms on PAM and BLOSUM Matrices
A large amount of new protein sequences is being accumulated rapidly in various databases. It is therefore important to develop fast and efficient methods to classify these proteins into their functional families. Many existing protein classification methods rely on multiple alignments. Generating reliable multiple alignments become problematic when dealing with extremely diverged protein sequences. In this study, several methods are compared that use multiple alignments with those that do not, and methods that use both negative and positive data for learning with those that only use positive data. Furthermore, there has been little study for using simply the amino acid composition for protein classification. Therefore, we examined the use of amino acid frequencies with various pattern recognition methods and compared their classification performance.