Han et al 2014 preset
alexell22 opened this issue · comments
Hi!
I am trying to analyse the data from the single sorted cells (compatible with the protocol Han et al 2024). However, I am getting Alignment failed: absent barcode: 3411672 (100%) when trying to run the preset. Is there a way to fix this issue?
Do you use the data set from the publication?
I managed to get the Han et al 2014 preset work on un-demultiplexed Undetermined_S0_L001_RN_001.fastq samples, but still got some uncertainties:
- We have 20*96 well plates coming from different samples and among the plate conditions are different. I am wondering if just mixcr analyze han-et-al-2014 --species hsa R1.fastq R2.fastq result would be enough to determine the precise location on the plate?
- I have also tried to add --sample-sheet samplesheet.tsv to have a more precise outcome, but I get Unmatched argument at index 8: 'output/' and the same with the result function (discussed in MIXCR Input file name expansion section).
the sample sheet is in tsv and has the structure
SAMPLE CELL3PLATE CELL1ROW CELL2COLUMN
sample1_A1 GCAGA TAAGC GTTCA
sample2_B1 GCAGA TGCAC GTTCA
sample3_C1 GCAGA CTCAG GTTCA
sample4_D1 GCAGA GGAAT GTTCA
...
(same structure for ~2000 rows)
Hi, so as I understand different plates are different samples?
Then you can indeed use the --sample-sheet, but there is no need to specify all othe barcodes, just the one that defines a sample (CELL3PLATE). I attach an example of the samplesheet.
The command should be:
mixcr analyze han-et-al-2014 \
--sample-sheet han-sample-table.txt \
--species hsa \
input_R1.fastq.gz \
input_R2.fastq.gz \
output
That way you will have different output files for each sample. If you have multiple plates per sample just use the same barcode twice in the samplesheet.
Sincerely,
Mark