Memory usage

[Bradley]

Hi Chris,

Thanks for the great paper and easy to follow analysis.

I want to repeat this but using all healthy samples (not just the 10 in the training set). I have tried using run_analysis.r but with all cells of each subset. This is fine until ComBat batch correction, at which point this fails due to a memory error, regardless of how much memory I request for my R session.

I have also tried running as a DelayedArray, but this still fails. Presumably this is because the data is still being read to disk by the ComBat function.

I was wondering if you could shed some light on how much memory you are using to run this analysis, or how you overcame this.

Thanks for your help,
Bradley

[Maarten de Vries]

Hi Bradley,
Were you able to resolve this?
I'm running into the same issue.

Thanks!
Maarten

[Bradley]

Hi Maarten,

Not using ComBat. I gave Harmony a go, but using reciprocal integration from Seurat gave me the best looking results batch-wise.

Best,
Bradley