We're seeing a particular VJ combo appearing in high numbers in many of our samples in the latest batches. This pattern was not observed in earlier batches. Is this a computational error or a wet bench contamination, or something else?
Work plan
Look for this combination in each sample in the two most recent batches (DNA160609LC and mammary_DNA160609LC) and evaluate how significant its presence is, (i.e. is it a dominant clone, or does it just have a few copies?)
Results
See analysis as well as attached files for full summary of each batch. Attached files: weird.vj.summary is for DNA160609LC and mammary_weird.vj.summary is for mammary_DNA160609LC
Conclusions
From the batch summaries it appears that there was some sort of contamination of our samples by the p14 samples. The spike count analysis does not suggest that the V13-3/J2-4 combination is particularly effective in amplifying, relative to other combinations. This is also observed in Burcu's primer independence file. In the clustered heat map, we see that V13-3/J2-4 is does not have counts significantly greater than expected.
I think we should delete this entry from all of the export_clone and export_align files before using them for analyses such as the standard pipeline analysis and/or the MiXCR QC analysis.
weird.vj.summary.txt mammary.weird.vj.summary.txt
Summary
We're seeing a particular VJ combo appearing in high numbers in many of our samples in the latest batches. This pattern was not observed in earlier batches. Is this a computational error or a wet bench contamination, or something else?
Work plan
Look for this combination in each sample in the two most recent batches (DNA160609LC and mammary_DNA160609LC) and evaluate how significant its presence is, (i.e. is it a dominant clone, or does it just have a few copies?)
Results
See analysis as well as attached files for full summary of each batch. Attached files: weird.vj.summary is for DNA160609LC and mammary_weird.vj.summary is for mammary_DNA160609LC
Conclusions
From the batch summaries it appears that there was some sort of contamination of our samples by the p14 samples. The spike count analysis does not suggest that the V13-3/J2-4 combination is particularly effective in amplifying, relative to other combinations. This is also observed in Burcu's primer independence file. In the clustered heat map, we see that V13-3/J2-4 is does not have counts significantly greater than expected.
I think we should delete this entry from all of the export_clone and export_align files before using them for analyses such as the standard pipeline analysis and/or the MiXCR QC analysis.