The HMS-LINCS Center is addressing a range of technical and conceptual challenges related to collecting and analyzing feature-rich data on cellular pathways and mechanisms of drug action. We look to collaborators and the broader scientific community to help us address these issues.

A significant but interesting challenge is determining how to link biochemical and imaging data on immediate-early cell signaling pathways (which typically involve changes in the levels, activities, and states of modification of receptors, kinases, and transcription factors) to data on other readouts being collected by collaborators at the other LINCS centers. Relatively few attempts have been made to link together different classes of data in a systematic and large-scale manner.

Data integration requires defining and instantiating new metadata standards and data ontologies and new means to organize the data we are collecting (e.g. XML-HDF5 file format SD-Cubes). Standards need to be drawn from and harmonized with existing community efforts and should meet the needs of both serious and casual users. We anticipate that it will take several years to develop a full-featured software pipeline, and we will therefore release data in provisional formats in the meantime.

Because the data we are collecting are heterogeneous, it will be informative to determine when they are in concordance or disagreement. For example, will the activities of kinase inhibitors measured in vitro on recombinant kinases correlate well with activities measured in cells? Across a panel of cell lines will the sensitivities of cells to inhibition of receptor tyrosine kinases correlate with their responsiveness to ligands that bind these receptors or to the level of receptors on cells? Across the same lines will total receptor levels correlate with levels on the cell surface or with transcript levels (or with complex gene signatures)? Our studies will hopefully answer these and similar questions.