virajbdeshpande/AmpliconArchitect

How can I run PrepareAA with tumor only sample(WGS)?

Opened this issue · 14 comments

edceeyuchen commented

I have some samples -only tumour, no any normal samples as control, and here is my code for these pipeline, it's OK?

python3 /share/home/luoylLab/zengyuchen/biosoft/AmpliconSuite-pipeline/PrepareAA.py -s ${tumor} -t 10 --ref GRCh38 --sorted_bam ${tumor_bam} --cnvkit_dir ${cnv} -o ${savedir}/${tumor}

jluebeck commented

Looks good - you may want to set --run_AA --run_AC if you are planning to launch AA and AC after.

edceeyuchen commented

Looks good - you may want to set --run_AA --run_AC if you are planning to launch AA and AC after.

Thanky for your quick reply! And I try this pipeline, find most samples have ecDNA(>50%), I worried about the high proportion, maybe there is better pipeline for tumor only samples calling AA?

jluebeck commented

For some cancer types (e.g. GBM) this rate is expected. I would be happy to look at any outputs or log files you would like to share to check for correctness of running the pipeline. AA .png images and AmpliconSuite log files are quick for me to assess and debug. My email is jluebeck [at] ucsd.edu.

Thanks,
Jens

edceeyuchen commented

Sure! I will send these files to your email! Thank you for your help!

edceeyuchen commented

For some cancer types (e.g. GBM) this rate is expected. I would be happy to look at any outputs or log files you would like to share to check for correctness of running the pipeline. AA .png images and AmpliconSuite log files are quick for me to assess and debug. My email is jluebeck [at] ucsd.edu.

Thanks, Jens

Sorry, I can't send my email to you(jluebeck@ucsd.edu) because of 'the message was rejected by the recipient'.

jluebeck commented

Strange - it is a Google-managed account and is able to get other messages. Perhaps can you try sending from a different email? It's also possible it does not like whatever format the attached files were in.

edceeyuchen commented

jluebeck@ucsd.edu

I has changed another email, and successed!

edceeyuchen commented

Strange - it is a Google-managed account and is able to get other messages. Perhaps can you try sending from a different email? It's also possible it does not like whatever format the attached files were in.

Hi jluebeck! We have another problem, when I do this pipeline for normal samples, there are ecDNAs also, it's so confused.

jluebeck commented

Thanks for sending the tumor-only files. I reviewed the files and the outputs look to be in order. I did not see any obvious bugs. In a small number of samples with previously established CN alterations (e.g. chromothripsis), 50% ecDNA is not a surprising number to me. However, I recommend though that you update your versions of AmpliconSuite-pipeline, AmpliconArchitect and AmpliconClassifier to the latest versions (e.g. AA is now at 1.3.r6 on the AmpliconSuite fork).

That is very unusual to see ecDNA calls from the pipeline on normal samples. Please feel free to also send the results for your normal samples and I will be happy to review. First though, I recommend that you check that you have the latest versions of all tools and re-run to see if you still obtain ecDNA+ calls in normal. It also helps if you are able to share the stdout printed by PrepareAA.py (now called AmpliconSuite-pipeline.py).

edceeyuchen commented

Thanks for sending the tumor-only files. I reviewed the files and the outputs look to be in order. I did not see any obvious bugs. In a small number of samples with previously established CN alterations (e.g. chromothripsis), 50% ecDNA is not a surprising number to me. However, I recommend though that you update your versions of AmpliconSuite-pipeline, AmpliconArchitect and AmpliconClassifier to the latest versions (e.g. AA is now at 1.3.r6 on the AmpliconSuite fork).

That is very unusual to see ecDNA calls from the pipeline on normal samples. Please feel free to also send the results for your normal samples and I will be happy to review. First though, I recommend that you check that you have the latest versions of all tools and re-run to see if you still obtain ecDNA+ calls in normal. It also helps if you are able to share the stdout printed by PrepareAA.py (now called AmpliconSuite-pipeline.py).

Very thanks for your quick reviewed. I run PrepareAA with the --sorted_bam ${normal} and not set the --normal_bam parameters, it's alright? And as you said, I will try to update my AA version from 1.3.r1 to 1.3.r6.

jluebeck commented

Yes, that should be an acceptable control. The files previously sent did not indicate a version of AmpliconSuite-pipeline, but please consider updating to latest (1.0.0): https://github.com/AmpliconSuite/AmpliconSuite-pipeline

edceeyuchen commented

Yes, that should be an acceptable control. The files previously sent did not indicate a version of AmpliconSuite-pipeline, but please consider updating to latest (1.0.0): https://github.com/AmpliconSuite/AmpliconSuite-pipeline

OK! I will try it and feedback as soon! Thank you again

edceeyuchen commented
edceeyuchen commented

Yes, that should be an acceptable control. The files previously sent did not indicate a version of AmpliconSuite-pipeline, but please consider updating to latest (1.0.0): https://github.com/AmpliconSuite/AmpliconSuite-pipeline

Hello Jens!
The germline samples results about AA had sent to your email! (and it's your google email),
Look forward to your suggestion and reply.