Error Running ASCAT

[Krishna4369]

Hi, I am getting below message from ASCAT Warning message: In runASCAT(lrr, baf, lrrsegm, bafsegm, ASCATobj$gender[arraynr], : ASCAT could not find an optimal ploidy and cellularity value for sample Tumor_LogR.txt

I'am running ASCAT for tumor sample only. I have attached input files and the Graph data generated by ASCAT.

Please suggest what could be the reason behind this....

Tumor_BAF.txt

[tlesluyes]

Hi @Krishna4369,

It's not really an error, it's a warning message saying that no good purity/ploidy fit was found for this sample. In more detail, ASCAT is leveraging information from heterozygous SNPs and this usually comes from matched germlines. When no germline is provided, ASCAT tries inferring which SNPs are heterozygous but such a process is noisier compared to providing true germline information. Therefore, purity/ploidy estimates are less accurate and, for some samples, no solid fit can be found.

Can you please provide all of the command lines that you have used? Which ASCAT version are you using? Which platform is it? Are you correcting logR for GC content and replication timing effect? Please also provide the plot coming from ascat.predictGermlineGenotypes (likely called tumorSep2200101233.png).

I was able to process this case on my side with the latest ASCAT version (v3.0.0; platform = "AffySNP6", although this needs to be adapted to your data; no logR correction). Perhaps updating ASCAT will help? I'm not sure how profiles (the last two plots) were generated if you got this warning message as no profile should have been generated at all. Did you re-run ASCAT with specific purity/ploidy values afterwards?

Cheers,

Tom.

[Krishna4369]

Hi,

I am extremely sorry for posting the wrong data in the forum. I was running ASCAT for 5 different samples and did not get a result for two samples. The one which I have posted (2200101233) and the other (2200103023). Then after checking all the parameters I resolved the error in the sample I have posted (2200101233) but unable to solve warning message/error message which i have posted in other sample (2200103023)

Here, I'm attaching the corrected BAF and LogR file. Please consider these files.

I'm using ASCAT version 2.5.2 on x86_64-pc-linux-gnu. I'm not correcting GC content for logR data and I've also attached the graph coming from ascat.predictGermlineGenotypes Tumor_LogR.txt Tumor_BAF.txt

[tlesluyes]

Hi @Krishna4369,

I still don't know what command lines you have used and what platform we're talking about so I can't really do anything. At that stage, my advice would be to try ASCAT v3.0.0 and correct logR for covariates (https://github.com/VanLoo-lab/ascat/tree/master/LogRcorrection). If ASCAT is still unable to find a purity/ploidy fit, then my recommendation would be to exclude this sample.

What was the tweak for fixing 2200101233?

Cheers,

Tom.

[Krishna4369]

Hi. Thank you for your inputs.

I have run the following commands ascat.plotRawData(rep_1_ascat.bc) ascat.gg = ascat.predictGermlineGenotypes(rep_1_ascat.bc, "AffyOncoScan") rep_1_ascat.bc = ascat.aspcf(rep_1_ascat.bc,ascat.gg=ascat.gg) ascat.plotSegmentedData(rep_1_ascat.bc) rep_1_ascat.output = ascat.runAscat(rep_1_ascat.bc)

For 2200101233 I have checked the data carefully and the data was having duplicate entry for a few of markers. I resolved that it worked.

[tlesluyes]

Hi @Krishna4369,

Looks good, I guess you have tried to remove duplicated entries in 2200103023 (like you did for 2200101233) but it didn't help, right? I would try correcting logR for covariates as well as running ascat.aspcf with penalty=70 (default value for ASCAT v3.0). If it still doesn't work, you'd have to exclude this sample.

Cheers,

Tom.