University of New Orleans bioinformatics group

Distributed toggle clusters learning

New! Here we implement distributed toggle clustering and classification of nanopore DNA hairpin toggles based on mixtures of HMM profiles. This approach is discussed in our recently submitted paper: Alexander Churbanov and Stephen Winters-Hilt, Clustering ionic flow blockade toggles with Mixture of HMMs, BMC Bioinformatics, 2008, MCBIOS V proceedings.
  • Here is the preliminary report on our approach slides
  • Corrected Figure 4 from the manuscript
  • Demo program

Duration learning demo

New! This program demonstrates the duration learning for the nanopore ionic flow blockade signal as described in our recent article Alexander Churbanov, Carl Baribault and Stephen Winters-Hilt, Duration learning for analysis of nanopore ionic current blockades, BMC Bioinformatics, 8(2007), Suppl. 7.
  • The corect picture for the profile shown on page 6 in Figure 4 could be seen here.
  • Demo program

Linear memory HMM demo

New! We managed to improve originally proposed linear memory implementation of Baum-Welch and Viterbi decoding for HMM as discussed in article by István Miklós and Irmtraud M. Meyer.

CSB2007 PM1 tutorial session

Here are the lecture slides describing the concepts of probabilistic modelling we use in our research.

Lecture on splicing

Here are the lecture slides describing the concept of Homology based gene structural prediction, Ab initio gene finding and splice site sensor design program.

Lecture on linear memory HMM

Here are the slides on our linear memory HMM implementation experiments.

Progress report on distributed toggle clustering with MHMM

Here are the slides on clustering with Mixture of HMM profiles. We have already run the algorithm in distributed manner using Java RMI.