Chronos Combined showing weird artifact not present in the individual datasets

Issues and Bugs Report an Issue
1

Hi,

I noticed a weird effect when looking at the various chronos datasets. In general, each independent dataset (Score OR Achilles Chronos vs Score+Achilles Chronos) seem pretty similar.

However, when you look at tumor suppressors that have a tendency to accelerate cell growth in the individual screens (TP53, NF1, PTEN), the increased growth rate looks like it has some weird artifact only in the combined Achilles/Score Chronos dataset, almost like it’s capped at a certain value.

For an obvious visual example :

image
image881×792 53.3 KB

I’ve read the Chronos paper a few times but am having trouble understanding why you would get this in the combined dataset. Is this something you guys have noticed before? Any ideas what’s going on here?

Thanks,
Dylan

2

Hi Dylan,

Thanks for pointing this out. This artifact is produced by the integration of the Achilles and Score datasets rather than Chronos. We’ll look into it further.

1 Like
3

Hi Joshua,

I have noticed this weird artifact as well and was glad to see that someone had already pointed it out. (Thanks, Dylan!) I was wondering if there are any updates on this issue.

Also, in the meantime, I was wondering what additional data processing/normalization is done during the data integration. Is it valid for me to download the Achilles and Score datasets separately for TP53 and then just combine the two datasets myself, and plot them on the same plot? Or is there additional normalization done during the data integration, which would mean it would be incorrect to plot the two datasets on the same scale?

Thanks so much!
Pia