We offer a comprehensive analysis service for all types of microarray.  The starting point is often the generation of lists of differentially expressed genes between two conditions in, say, an Affymetrix GeneChip experiment.  From this most basic level, we can then go on and produce more in-depth results.  For some ideas of what is possible, please look over the following sections - but if you have any other requirements, feel free to come and discuss them with us.

Complex Designs

A large proportion of experiments are not just looking at the difference beween two conditions, but might have multiple experimental groups (for example, treated and untreated conditions in two cell-lines, making four conditions in all).  There are many different questions answerable in such situations (which genes respond to treatment in cell-line A but not in cell-line B; which respond in both, but to different degrees; ...), and we have techniques available for generating such genelists.  Similarly, when a time-series experiment has been carried out, it's often not quite so simple to formulate exactly what characterises "interesting" genes: we have plenty of experience that you can call upon for analysing such data.  You can also look at the biostatistics section to gain a bit more understanding of the issues involved.


Beyond the Genelist

Large genelists can be bewildering objects, and beyond validating a few top targets, what can be done to help understand them?  We can provide a preliminary biological explanation of the contents of your genelists in terms of Gene Ontology, attempting to capture associated biological processes, cellular components and molecular functions that are over-represented in you experiment; additionally, we can look to see if genelists from an external source are acting in specific ways within your experiment.  We can also pull out empirical clusters of genes that appear to be operating in a related manner.  Beyond this, we can carry out pathway analysis, or if you have specific requirements to add biological information onto the genelists, do come and discuss your ideas.


Diverse Technologies

We don't just deal with Affymetrix experiments, nor confine our work to expression analysis.  We endeavour to help with all data from CRUK scientists, such as
  • SNP arrays deserve their own section
  • General expression arrays - we can usually apply the same methods we use for Affymetrix analysis to any other expression technology, including Illumina, Applied Biosystems, two-colour spotted arrays, ...
  • Array CGH - we can: help identify breakpoints in copy number in individual samples; find clones that are have copy number variation that correlates with clinical data across samples; find regions of common loss/gain across clinical samples; find pairs of clones that appear to be acting in a correlated manner across samples; ...
  • Tiling arrays - beyond the types of technique that are applicable to aCGH above, we can help identify peaks where ChIP-chip experiments might indicate binding.
  • Methylation arrays, Protein-protein interaction arrays, ...