visualizing sgRNA-response pairs with Nextflow output

[nlthanhthanh]

Hello -- big fan of sceptre here!

I successfully ran Nextflow Sceptre through to the final steps of association analysis. I was wondering if there is any way to produce sgRNA-response pairs plots similarly to plot_response_grna_target_pair function using Nextflow outputs (which include gRNA assignment matrix and .RDS files with summary statistics). Many thanks!

Thanh Thanh

[timothy-barry]

Hi, good to hear!

Would it be feasible for you to run the gRNA assignment and QC steps within the R console? If so, my recommendation would be to run the pipeline out-of-core up through run_qc() within the R console (according to these instructions). Then, you should be able to call plot_response_grna_target_pair() on the sceptre_object. Note that the p-value will not be shown; you will have to superimpose this yourself. I think this should work, but please let me know if you hit any snags.

[nlthanhthanh]

Thank you for your response! My dataset is big so it wasn't feasible to run out-of-core. However, I dug through the Nextflow cache files and were able to locate the sceptre_object at the run_qc step!

Which unfortunately brought up a different problem: Because I used bonferroni as sgRNA integration strategy, plot_response_grna_target_pair gave me error "gRNA target is not present" due to the grna_group_idxs within sceptre_object contains gRNA IDs as singletons. Would you mind explain (1) how the p-values of gRNA IDs with the same targets were combined into 1 single p-value with bonferroni? and (2) How should I then plot this accordingly (e.g., was the gRNA ID with the lowest corrected p-value the final result) ?

[timothy-barry]

Hi,

You can read about the bonferroni gRNA integration strategy here. We need to improve the interface to plot_response_grna_target_pair() when grna_integration_strategy is bonferroni, but yes, my recommendation would be to select the gRNA with the lowest p-value for the plot.