Challenges in Bioinformatics

1
Challenges in Bioinformatics R.W.
Doerge Department of Statistics Department
Agronomy Purdue University
2
A second code of instruction DNA
modifications
Methylation profiling
3
Reflections

• What is bioinformatics?
• Bioinformatics is the application of computer
  technology to the management of biological
  information
• Computers are used to gather, store, analyze and
  integrate biological and genetic information
  which can then be applied to gene-based drug
  discovery and development
• The science of Bioinformatics, is the melding of
  molecular biology with computer science
• Universities, government institutions and
  pharmaceutical firms have formed bioinformatics
  groups to unraveling the mass of information

4

• Technologies evolve
• More and more data

Cell 157, March 27, 2014
5
Looking Forward

• Bioinformatics uses many areas of computer
  science, statistics, mathematics and engineering
  to process biological data
• Not just any data, the right data
• and, a lot of it

6
Looking Forward

• Data storage
• one human genome (DNA) 200 gigabytes
• five human genomes 1 terabyte
• Data access
• Data analysis

7

• Principal Component Analysis reduces
  multidimensional (single cell) data by
  identifying linear combinations of genes that are
  responsible for cell-to-cell variability
• measurement of mRNA distributions allows
  determination of kinetic parameters controlling
  expression of individual genes
• (top) Slow transitions between the on and
  off state of a promoter can give rise to
  bimodality
• (bottom) Fast transitions lead to unimodal copy
  number distributions
• Genes (left) controlled by the same upstream
  regulator are expected to be positively or
  negatively correlated across single cells
• Clusters (right) of co-regulated genes identified
  via pairwise correlations

Cell 157, March 27, 2014
8
Looking Forward

• Personalized medicine
• Personalized food
• Mechanisms underlying heterogeneous gene
  expression
• transcription factor binding
• methylation
• histone modifications
• single cell nucleosome occupancy
• spatial orientation of single cells in tissue

9
AHEAD, Nature 2008
10
Opportunities

• an epigenome exits
• per cell?
• it is dynamic
• epigenomic changes control gene expression
• epigenomic marks are heritable

11
Opportunities

• Pre-processing of emerging high throughput data
• Dependence in high-dimensional data
• high dimensional discrete counts
• Integration of multi-omics data
• Modeling dynamics of mixtures
• populations of cells, variants, metagenomics
• Big data approaches for addressing 'omics'

12
Opportunities

• 10-70 trillion cells in the human body
• genome per cell
• epigenome per cell (type)?
• variation between individuals, tissues, cells,
  across time
• epigenome is dynamic
• interactions between the environment, the
  epigenome, and genome
• Think of the opportunities

13
(No Transcript)