Bioin401 Project: Microarray Pipeline

1
Bioin401 Project Microarray Pipeline

• Chunyan Meng,
• Yifeng Liu,
• Chelsea Ju
• March 27, 2006

2
Microarray Pipeline GANTT Chart
3
Micorarry PipelineProgramming progress

• Bio-Perl installation and testing
• Email notification

4
Hiearchical Clustering Gene function prediction

• Checked with Greg about biospider
• Checked with Savita about BASys
• (Bacterial Annotation System)
• Literature review about gene function
  prediction
• Browser the current available methods used for
  gene/protein function prediction
• Trying to find the software or paper using human
  microarray data

5
Microarray PipelineGene Function Prediction
software
6
Microarray PipelineGene Function Prediction
7
Microarray PipelineGene Function Prediction
(STRING)

• STRING - Search Tool for the Retrieval of
  Interacting Genes/Proteins
• STRING is a database of known and predicted
  protein-protein interactions. The interactions
  include direct (physical) and indirect
  (functional) associations they are derived from
  four sources
• Genomic Context analysis conserved genomic
  neighborhood, gene fusion events, and
  co-occurrence of genes across genomes.
• High-throughput Experiments microarray
  experiments 
• (Conserved) Coexpression 
• Previous Knowledge (text mining Pubmed and mips)
•     

8
Microarray PipelineGene Function Prediction
(STRING)

• STRING quantitatively integrates interaction data
  from these sources for a large number of
  organisms, and transfers information between
  these organisms where applicable.
• The database currently contains 736429 proteins
  in 179 species.
• Input gene/protein name or amino-acid sequence
• Scope 179 species
• Access publicly accessible through web
  interface, academic license free but with delay.
  Selected data-items as flat files are available
  for free.

9
Microarray PipelineGene Function
Prediction(Phydbac)

• Phylogenomic Display of Bacterial Genes (Phydbac)
• The non-homology based methods allow to determine
  proteins potentially linked to the pasted
  sequence. Functional predictions are then made
  out of the annotations of the proteins associated
  with the query.
• The comparative genomic methods used here are
  - Phylogenetic Profiling identifies proteins
  that have evolved in a similar manner to the
  sequence of interest. - Genomic proximity
  identifies proteins that are found nearby on 1 or
  more genome.- Domain Fusion events determines
  the Pfam domains that are found fused with
  domains matching the given sequence.
• Input a gene name or sequence in fasta format
• Scope bacteria E Coli K12
• Access publicly available through web interface

10
Microarray PipelineGene Function Prediction

• Get better understanding what function prediction
  is all about. What kind of methods are available.
  What kind of challenges exist.

11
Microarray PipelineWhat to do next

• Paper review on gene function prediction
• Integrating different parts together
• Project report writing

12
Microarray Pipeline GANTT Chart
13

• Any comments?