Applications of generalized pair hidden Markov models to alignment and gene finding problems

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• Hidden Markov Models (HMMs) are widely used in molecular biology for tasks like gene finding and sequence alignment.
• Pair HMMs solve alignment problems, while generalized HMMs model exon lengths for gene finding.
• Existing methods lack a unified framework for complex cross-species analyses.

Purpose of the Study:

• Introduce the generalized pair HMM (GPHMM) as an extension of existing HMMs.
• Demonstrate the utility of GPHMMs for cross-species gene finding.
• Provide a unified and probabilistically sound model for DNA-cDNA and DNA-protein alignment.

Main Methods:

• Developed the generalized pair HMM (GPHMM) by extending pair HMMs and generalized HMMs.
• Utilized approximate alignments in conjunction with GPHMMs for gene finding.
• Applied GPHMMs to DNA-cDNA and DNA-protein alignment problems.

Main Results:

• The GPHMM framework effectively integrates features of both pair and generalized HMMs.
• Demonstrated successful application of GPHMMs for cross-species gene finding.
• Showcased the model's utility in DNA-cDNA and DNA-protein sequence alignment.

Conclusions:

• GPHMMs offer a powerful and unified approach for diverse molecular biology problems.
• This model advances the accuracy and scope of cross-species gene finding.
• GPHMMs provide a robust theoretical foundation for sequence alignment and annotation.