search

aeeckhou / shallowHRD

This method uses shallow Whole Genome Sequencing (sWGS) and the segmentation of a genomic profile to assess the Homologous Recombination Deficiency of a tumor based on the number of Large-scale Genomic Alterations (LGAs).
30 stars 13 forks source link

Why are some LGAs not registered #17

Closed carolinehey closed 10 months ago

carolinehey commented 10 months ago

Hello again, I apologize for my many questions, but I am very keen on integrating shallowHRD into our tumor sequencing workflow. I have run this sample, but I wonder why the LGAs I see in plot A including chromosome 5 and 8 are not registered. I hope you can inform me.

G37-T42_illumina-truseq-genome_H2H7JDSXX-H2HCKDSXX_filtered_summary_plot

Best, Caroline

aeeckhou commented 10 months ago

Hello Caroline,

You shouldn't assess the existence of a LGA based only on the visual ! Here I would look for the final segmentation, how it is (Files "final_segmentatio.jpeg" & "final_segmentation.txt").

Did my segment that I believe are composing the LGA are there, at the right ratio ? Then look at the size of the segment. Maybe one segment is only 9Mb long i.e. it is too short to be counted as an LGA. or maybe the distance between them is too important to be counted as an LGA. Finally, look at the difference between the two segments ratios in terms of ratio value. Is it below the CNA cut-off value ? If so they should be counted as a LGA and might be subclonal but not detected with the CNA cut-off value.

I can't really answer with only this final output sorry.

Best, Alexandre