Average Contribution Score

[shunyasanuma]

Hello,

I have a question regarding the average DeepLIFT/contribution score.

Let’s assume you have multiple repeats of GATA, where the first GATA has scores of 0.8, 0.8, 0.7, and 0.8. The second GATA also has scores of 0.8, 0.8, 0.7, and 0.8. However, if the third GATA has scores of 0.01, 0.02, 0.01, and 0.01 (much smaller than the others), the average score decreases and may not reflect the correct biological significance (it could cause underestimation).

Does tfmodisco use thresholds to filter out cases like the third GATA in the example above?

[AvantiShri]

Hi Shunya, technically yes in the sense that if the third GATA has very low scores, it would not get through the seqlet identification step and thus would not make it through to the subsequent clustering steps. Hope this helps.

…

On Thu, 18 Jul, 2024, 03:29 Shunya Asanuma, ***@***.***> wrote: Hello, I have a question regarding the average DeepLIFT/contribution score. Let’s assume you have multiple repeats of GATA, where the first GATA has scores of 0.8, 0.8, 0.7, and 0.8. The second GATA also has scores of 0.8, 0.8, 0.7, and 0.8. However, if the third GATA has scores of 0.01, 0.02, 0.01, and 0.01 (much smaller than the others), the average score decreases and may not reflect the correct biological significance (it could cause underestimation). Does tfmodisco use thresholds to filter out cases like the third GATA in the example above? — Reply to this email directly, view it on GitHub <#116>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AARSFBQCYGCLYI5J6IIAK3LZM3SL5AVCNFSM6AAAAABLBOWQM6VHI2DSMVQWIX3LMV43ASLTON2WKOZSGQYTINRWHE3TQMA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>