This pull requests fixes some typos in the 10x processing code.
Additionally, it concatenates the I1 reads to the front of the barcode (since these are sample barcodes), and then calls from 0 - 24 the barcode (reflecting 0-8 the 8 bp sample barcode, and 8 - 24 the 16 bp cellular barcode), and 24-34 the UMI (reflecting the 10 bp UMI barcode). see #2 for explanation.
I believe that this may help with #3 as well, since the current method treats the sample barcode as the UMI and hence in a single sample there may be only 4 barcodes, leading to problems with downstream UMI analysis.
This pull requests fixes some typos in the 10x processing code. Additionally, it concatenates the I1 reads to the front of the barcode (since these are sample barcodes), and then calls from 0 - 24 the
barcode(reflecting 0-8 the 8 bp sample barcode, and 8 - 24 the 16 bp cellular barcode), and 24-34 the UMI (reflecting the 10 bp UMI barcode). see #2 for explanation.I believe that this may help with #3 as well, since the current method treats the sample barcode as the UMI and hence in a single sample there may be only 4 barcodes, leading to problems with downstream UMI analysis.