Han el all preset

[alexell22]

Hi!

I am using the Han et all preset (han-et-al-2014) , with mixcr-4.3. I am facing several issues:

1. Low coverage of the plates (I have 20 plates in total, 5 sorted cells in each well, library prep is done the same way as discussed in the paper, and currently libraries are resequenced). However, I am getting quite a similar outcome with the specified preset (~50%). What could be the reason behind it, if not the library prep?
2. Some Plate barcodes could not be found in the processed data (plate is missing), however are present in the unprocessed data (with the required seq, as well in the ROW and Column). Is it possible to check the builtin:han-2014-cell3-plate? Or what can be also done in that case?

Thank you!

[mizraelson]

Hi,

1. The preset is optimized to work with datasets with one cell per well as was done in the original publication.
2. Can you share the report files?
3. I am attaching the list of the plate barcodes build into the preset.
   han-2014-cell3-plate.txt

[alexell22]

Sure!
result.align.report.txt
result.assemble.report.txt
result.qc.txt
result.refine.report.txt

Thank you for the list!

[mizraelson]

Did you use the same primers as in the paper?

[alexell22]

Yes, the same (1st, 2nd reaction and barcoding). We did not include the phenotyping ones, just for TCRs.

[mizraelson]

From the reports, I see you are using MiXCR v4.6 not 4.3

Here is the pattern from the preset:

 ^NN(CELL3PLATE:N{5})ga(CELL1ROW:N{5})ccagggttttcccagtcacgac(R1:*) \
 ^NN(CELL2COLUMN:N{5})gtcactggatttagagtctctcag(R2:*) |
 ^NN(CELL2COLUMN:N{5})gagatctctgcttctgatggctc(R2:*)

According to the reports, the main issue is that this pattern wasn't matched:

Alignment failed: absent barcode: 4197136 (94.79%)

Do you see these primer sequences in your data? If you can share the FASTQ files, I can take a look.

[mizraelson]

The drive link would work. You can send it to [email protected] and I will take a look.

[alexell22]

Thank you, have shared the files! :)

[mizraelson]

Hi, is it intentional that you have all reads as Undetermined? I checked the FASTQ file and most reads (>70%) align to PhiX which is expected for the Undetermined FASTQ file. The rest does not really seem to have the primers sequences in it even if it aligns to TCR.

For example:

@M01338:7:000000000-LL5J2:1:1101:15600:2226 2:N:0:0
TCGCAACACCTTTTTCAGGTCCTCTGTCACCGTGAGCCGTGTGCCCGGCCCGAAGTACTGCTCGTTCGCCCGCCCCTCCTCAAGCCAACCAATTCAGCTTTGGAACCACCTCTCTCCAGTCTCTTCAATACTTCCTGTCAGACTGGACAAGGTAGGGCTTGCATGCTTGGTGGGCAACTTGTCCTTCTCGACACCTTGCTCGTATGTGGCCTTGGAGCCCTCATTGGAAGTTGCCATCAGCATGAGACTCTGTTTCGGGAACTGAAGATACCAAAACATAGTTTTAGCCTGAAATTACAGA
+
-AC-,@FGG,FG,EFGGGGGGGGGG86,CF,BFGGGGG78CG,EF,CGGGGGGGG6CFGCGGG,+,,,+8C+++++,,,,,,,,:F,E,B4,,,C,C?,,C,,,C,A,+4,,CF,,A,<,,C,,,,,45,,,,8,8,4,,,4,4@,,6@E,,3,,+>@E8+363C,@,=+@F+++,3@CCCEG7CF,,7*;9**4;7=D+;?;DF49+/CFFGA4*)22:66)+=64*80AFEF*8<A=*4+4*/4433A@EEEFF:0(1:60)-))..-60348:?<<42::)-)(-5><0,2)6))64)

The read aligns to TCR:

But, the structure MiXCR is looking for in R2 is:

^NN(CELL2COLUMN:N{5})gtcactggatttagagtctctcag(R2:) or ^NN(CELL2COLUMN:N{5})gagatctctgcttctgatggctc(R2:)

Thus it is not clear where is the expected CELL2COLUMN barcode.

[alexell22]

Hello! Thank you for the response! This is something that we got from the core facility after sequencing. Okay, I see, is it possible to drop the CELL2COLUMN in the preset? Would it be also possible to have an online meeting with you for the discussion?

[mizraelson]

Of course, I will forward the meeting details via email.