Motifs, profiles and hidden Markov models

11/10/02


Click here to start


Table of Contents

Motifs, profiles and hidden Markov models

The objects of our study

The statistics of biological sequences can be global or local

Motifs - Sites - Signals - Domains

Specific searches / predictions

Motifs and models

PPT Slide

Prosite patterns

Consensus in sequences

Example: C2H2 zinc finger DNA binding domain

Searching with regular expressions

Regular expressions can be limiting

Cys-Cys-His-His profile: sequence logo form

Calculation of a position-specific scoring matrix (PSSM) from counts

Derivation of PSSM entries

Use of a PSSM to find sites

Representation of motifs: the next steps

Profiles

Derivation of a profile for Ig domains

PPT Slide

Hidden Markov models

HMMs: generalities

PPT Slide

Hidden Markov models:extensions

PPT Slide

PPT Slide

A very short profile HMM

How profile HMMs work: in brief

Pfam domain-HMMs

PPT Slide

PPT Slide

Designing the HMM, I

Designing the HMM, II

Designing the HMM, III

HMM: decoding

CC-PROBABILITY PROFILE

References

HMM-type software available

References

Author: WEHI ITS