Oberwolfach Seminar: Bioinformatics Approaches for Finding cis-regulatory Motifs and Modules

Date:

October 9th - 15th, 2005

Organizers:

Wolfgang Huber, Cambridge, UK

Xiaole Shirley Liu, Cambridge, Mass.

Terry Speed, Melbourne/Berkeley

Program:

Although (almost) every human cell contains a complete genome, and so has the capacity, in principle, to express every gene, cells at different stages in human development, or in different tissues, or in the same tissue under different conditions, typically exhibit very different patterns of gene expression. The process leading to a gene being transcribed in a cell is a complex one, involving many proteins sending and receiving signals, which, under certain circumstances, lead to the initiation of the transcriptional machinery at the gene's transcription start site (TSS). This process involves one or more proteins called transcription factors (TFs), together with what are called co-factors, binding to the genome "near" the TSS, at places called binding sites (TFBSs). This aspect of the process is called cis-regulation and the proteins involved are called regulatory elements. The general theme of this seminar is finding these regulatory elements and the associated binding sites. In doing so, we make use of cis-regulatory motifs and modules (i.e., clusters of motifs), these being regions of the genome which interact with the transcription machinery. A variety of types of genomic data and a wide range of computational methods have been brought to bear on this problem, and this seminar will review them. The data includes gene expression microarray data, chromatin immunoprecipitation (ChIP) data, and genomic sequence data for one or more species. In this context, finding includes "detecting", when the nature of the motif is known, and "discovering", when the motif is not known, but needs to be inferred. When we speak of "finding" clusters of motifs, there are 3 options: detecting clusters of known motifs of known form, discovering novel clusters of known motifs, and discovering novel clusters of novel motifs. In carrying out the task, a wide range of statistical methods have been found to be fruitful, as has the biological notion of evolution, more specifically, of comparative genomics. Our approach will be to explain the biological problem and the technology leading to the data, and then turn to the mathematical, statistical and computational methods involved in doing the "finding". The statistical topics we plan to discuss in this context include:

• robust linear models for microarray data
• scan and related statistics
• pairwise and multiple DNA and protein sequence alignment
• probability models for motifs and modules (including HMMs)
• phylogenetic inference
• product multinomial and dictionary models for DNA sequence
• MCMC and Gibbs samplers
• linear models connecting sequence, expression and ChIP data

Lecture Program:

Monday
Theme: Background and motivation for the week.
AM

1. Introduction to molecular biology
2. Microarrays and low level analysis
3. Transcriptional regulation in eukaryotes

PM
Computer Lab: cDNA and Affymetrix microarray analysis Spellman et al, 1998. Mol Bio Cell. Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Cho et al, 1998. Mol Cell, A genome-wide transcriptional analysis of the mitotic cell cycle.

Tuesday
Theme: Motif and module representation and detection
AM

1. Introduction to motif representation and detection
2. Profile and other HMMs
3. Representation and detection of modules

PM
Computer lab: Motif-detection (as distinct from discovery) programs. Retrieve intergenic sequences, run web motif finders. Paper: Tompa et al. 2005 Nat Biotech. Assessing computational tools for the discovery of transcription factor binding sites.

Wednesday
Theme: Motif discovery - sequence-based methods
AM.

1. Methods using the EM-algorithm
2. Methods using Gibbs samplers
3. Methods for the discovery of modules

PM
Hiking

Thursday
Theme: Motif discovery - using expression and ChIP data
AM

1. Linear models to discover cis-regulatory modules
2. Analysis of ChIP-chip experiments
3. MDScan and Motif Regressor

PM
Computer lab: Revisiting the yeast cell-cycle data and yeast ChIP-chip data for RAP1. Illustration of algorithms from Wednesday and Thursday mornings.

Friday
Theme: Bringing everything together AM

1. Motif finding in higher eukaryotes: comparative genomics
2. Motif clusters and cis- regulatory modules
3. Human ChIP-chip on tiled arrays

PM
Discussion of papers: Wasserman et al. 2000 Nat Genetics. Human-mouse genome comparisons to locate regulatory sites. Carroll et al. 2005 Cell. Chromosome-Wide Mapping of Estrogen Receptor Binding Reveals Long-Range Regulation Requiring the Forkhead Protein FoxA1. Gupta and Liu. 2005 PNAS. De novo cis-regulatory module elicitation for eukaryotic genomes.

Prerequisites:

Participants should probably have a basic knowledge of the biology of gene expression and gene regulation, although this will be briefly reviewed. Also the basics of statistics and probability, including stochastic process will be assumed.

Deadline for applications:

September 1, 2005

The seminars take place at the Mathematisches Forschungsinstitut Oberwolfach. The number of participants is restricted to 24. The Institute covers accommodation and food. Travel expenses cannot be reimbursed. Applications including

• full name and address, including e-mail address
• present position, university
• name of supervisor of Ph.D. thesis
• a short summary of previous work and interest

should be sent as hard copy or by e-mail (.ps or .pdf file) to:

Prof. Dr. Gert-Martin Greuel
Universität Kaiserslautern
Fachbereich Mathematik
Erwin Schrödingerstr.
67663 Kaiserslautern, Germany
.