Disease related signatures and treatment/baseline conditions

[Hendrik Weisser]

Dear iLINCS team,

In your library of "Disease related signatures", a "Treatment" condition is compared to a "Baseline" condition. Often, these conditions are reversed from what one would expect (or at least it is presented that way in the web portal) - "Treatment" is "control" and "Baseline" is a treatment (e.g. http://www.ilincs.org/ilincs/signature/GDS_5711), or "Treatment" is an early timepoint and "Baseline" a late one (e.g. http://www.ilincs.org/ilincs/signature/GDS_6098).

In these cases, are the expression signatures and resulting Concordance (correlation) values simply the inverse (i.e. "-1 * x") of what they would be with the more intuitive (i.e. reversed) condition assignment?

(I am looking for disease signatures that are anti-correlated to expression changes from a drug perturbation, so it is important to interpret the "direction" correctly.)


Best regards,

Hendrik Weisser



This e-mail including any attachments is for the intended recipient(s) only and contains information or material, which may be confidential and/or legally privileged. If you are not the intended recipient, any disclosure, copying, use, distribution or retaining of the information or material contained in this e-mail is unauthorised and prohibited. Please notify the sender immediately by e-mail if you have received it by mistake and delete the e-mail immediately from your system. We have taken steps to ensure that this email (and any attachments) are free from computer viruses and the like. However, it is your responsibility to ensure that it is actually virus free. Any emails that you send to us may be monitored for the purposes of ascertaining whether the communication complies with the law and our policies. Storm Therapeutics Limited. Registered Office: Moneta Building, Babraham Research Campus, Cambridge, England, CB22 3AT. Registered in England under number 09553473.

[Mario Medvedovic]

Dear Hendrik,

You are correct about the ordering of the "Treatment" and "control" samples in "Disease related signatures" and the "direction" of the signatures. These signatures have been created by an automated script which did not attempt to recognize which group of samples would be a natural "control" in the comparison. The signatures are always constructed as log(Treatment) - log(Baseline). In situations such as the one that you described, the signature indeed needs to be multiplied by -1 to get the natural orientation.

Best regards,

Mario

---

Mario Medvedovic, iLINCS PI