Export data

This Wiki page is part of piLINCS quick help. This page deals with data export at different levels:

* Raw data: high level data from Panorama with individual measurements using the QC+NORM normalization
* Profiles: data reorganized as actual profiles with individual replicates
* Merged profiles: data reorganized as merged/collapsed profiles with replicates collapsed into averaged profiles

These profiles can be exported in conjunction with user defined filters and tuples (assay, cell type, perturbation). Just follows the steps below.

• STEP 1: go to piLINCS
• STEP 2: go to Export tab
• STEP 3: use 'Filters' in conjunction with a recommender system to restrict your set of profiles to a subset of tuples (assay, cell type, perturbation)

For STEP 3, just start typing the name of your favorite cell type (e.g., M to see MCF7 listed below) or perturbation (e.g. R to see Resveratrol listed below).

• STEP 4: use radio buttons of the 'Filters' to define the type of profiles to be exported, selecting from

  • Raw data: original profiles from Panorama, each replicate/data point as individual entry
  • Profiles: original data from Panorama, but organized as actual profiles
  • Merged profiles: replicated collapsed to average profiles (with min and max values) given

• STEP 5: use the Export button or use the corresponding API link included below, with "Specified export is also available with API call" prompt.

Note that the data exported in STEP4 will be a subset constructed in conjunction with tuple filters defined in STEP 3. For example, for (P100+GCP,MCF7,Resveratrol) tuple the following API call (and the corresponding data set) will be constructed (with %22 representing "):

http://www.eh3.uc.edu/pilincs/api-assays-paged/?order=asc&limit=100000&offset=0&tags=[{%22name%22:%22MCF7%22,%22flag%22:%22Cell%22,%22annotation%22:%22CellId%22},{%22name%22:%22Resveratrol%22,%22flag%22:%22Perturbation%22,%22annotation%22:%22Pertiname%22},{%22p100%22:true},{%22gcp%22:true}]

Note also, that the format used to export data through this mechanism is JSON, with a structure that reflects the type of profile (raw vs. profile vs. merged profile) and should be easy to decipher (search for entries within JSON nested lists delineated by {}).