Fionagh Thomson       04:46 PM
Great Digital Pathology casestudy.  you mentioned, your source clinical data (for training neural networks) is from John Hopkins (USA).  Which other datasets would you like to access but are unable to? (either because it’s not yet collated or you cant get access)?

Alexander Szalay       04:50 PM
Part of our data also came from Yale. We have access to data from Bristol-Myers-Squibb, and a constortium of a large NCI project. For the open cell atlas we would like to cover as much as we can from the TCGA (The Cancer Genomic Atlas) with about 4,000 samples. We already have the slides for melanoma, as well as the corresponsing genomic data. We expect to do this in about a year.