Adjusting the parameters to be more sensitive

[ysbioinfo]

Hi, Thanks for developing such an awesome tool! I'm now using viral-track on my 10X data. They are from tumors infected by EBV. However, EBV cannot be detected by viral-track in most of the samples. Certainly, it could be attributed to the low sensitivity and high drop rate of 10X technique, but I want to know could the result be improved (higher sensitivity) by adjusting some parameters like entropy?

Thanks!

Best Yang

[PierreBSC]

Dear Yang,

Thanks a lot for your interest in Viral-Track. I suppose you are sequencing Burkitt's lymphoma ? There might be several reasons why you do not observe EBV :

• The infection is latent : the amount of transcribed viral RNA is too low to be detected.
• The infection has been wiped out : I have seen that very frequently with SARS-CoV-2/COVID-19, as patients in ICU have generally eliminated most of the viral load.
• The tissue sampling is not optimal : the infection is spatially localised within the tumor and you only analyzed a region which is not infected.

To my knowledge EBV RNAs should be easily detected as they are polyA+ so I think that changing the parameter values will not improve anything... Have you checked the txt QC report file ? And concerning the positive samples how many EBV reads do you find ?

Hope this will help ! Best, Pierre

[ysbioinfo]

Hi Pierre, Apologize for the late reply as I was away on vacation. Thanks for your suggestions. Very make sense. My samples were from gastric cancer and they were confirmed as EBV positive by PCR. I think another reason for not detecting EBV maybe due to the high dropout rate of 10x Genomics (I used 10x 5' technique). Here I attach a QC report file of one sample. May be you can get some insight from this file? Thanks! QC_unfiltered.txt

Best Yang

[ysbioinfo]

BTW, It seems that the Expression_table.tsv is a virus-level expression. Is there some way to get a gene-level expression? For example, there are several genes for NC_007605 (Human gammaherpesvirus 4, a subtype of EBV), like ENBA1, ENBA2, EBER, etc. Do you have any suggestion on how to get the expression of individual gene?

Best Yang

[PierreBSC]

Dear Yang,

When you mention PCR, you mean regular DNA PCR or just qPCR ? I had a look at the file you sent me and basically Viral-Track was able to identify the virus : NC_007605 and NC_009334. However the amount of RNA detected is VERY low (~2000 reads, not even UMIs for each)... So I would definitely favour a natural low abundance hypothesis ! Hope this helps, Best, Pierre

[ysbioinfo]

Hi Pierre, We performed qPCR of EBER. I totally agree. Thanks for your help!

Best Yang

[PierreBSC]

Hi Yang, You are welcome ! I am closing the thread but if you need help do not hesitate ! Best Pierre