/hmm

Find the most likely annotations of a sequence

Primary LanguageJavaScriptMIT LicenseMIT

HMM

How to run

  • Install node
  • Clone this repository
  • Viterbi: run npm run start:viterbi
  • Posterior Decoding: run npm run start:posterior-decoding
  • Viterbi Training: run sh src/gene-hmm-train/viterbi-training-example.sh
  • Baum Welch: run sh src/gene-hmm-train/baum-welch-example.sh

This is a small project that:

  • Computes the most likely annotations of a sequence given an HMM (Viterbi algorithm).
  • Computes the posterior-decoding (using the Forward and Backward algorithms).
  • Inferring parameters for an HMM (using Viterbi training or Baum-Welch).

That project handled as an homework assignment for the course: "Algorithms in Computational Biology".