Wolfram Technology Conference 2010

Oct 13-15, 2010
Here is the abstract of my talk:

Using Mathematica to Evaluate the Effect of
Transcription Factor Binding at Enhancers
on Gene Expression in Humans

A deep understanding of variations in gene expression patterns is necessary to
account for the large range of phenotypic variation in the human population.
This knowledge could open up the possibility of artificially changing gene expression,
which in turn offers numerous applications, ranging from temporary increases in
the production of necessary proteins such as hemoglobins,
to the permanent inhibition of cytokines as another option for treating cancer.

In this pilot study, the relationship between transcription factor binding
at distal control elements known as enhancers and subsequent gene expression
was investigated using data from the 1000 Genomes Project.
Mathematica was used for scripting and multivariate analysis.

First, we confirmed that Single Nucleotide Polymorphisms (SNPs) are responsible
for differences in transcription factor binding at enhancers.  It was also found
that these differences in transcription factor binding have an effect only on genes
a certain distance (denoted as 'critical intervals') away rather than on the nearest
gene, contrary to previous literature that enhancer activity is distance-independent
(all p-values were lower than 0.05).  Finally, our project showed that groups of
enhancers and genes that demonstrate synteny, where two or more loci are
located next to each other for various species, are not necessarily linked,
despite previous belief (i.e. restricting by syntenic group failed to lower
any of the p-values).

In conclusion, our methodology can identify different critical intervals on a gene
to which different transcription factors can bind and influence expression.
This will facilitate coupling enhancers with their target genes, allowing
for further quantitative research on enhancer activity.