Clonotype file is being read in wrong - the same clonotype is being split into 2 different clonotypes.

[ms3389]

🐛 Bug

Clonotype file is being read in wrong - the same clonotype is being split into 2 different clonotypes.

To Reproduce

Steps to reproduce the behavior:

1. If I load up this clonotype file from 10x sequencing: Patient 6 M6.csv

2. Immunarch splits a single clonotype with 3 CD3.aa sequences into 2 separate clonotypes.

3. lines 2 and 6 in the screenshot. It's the same 3 CD3.aa sequences (same nt sequences as well) just in different order

Expected behavior

Lines 2 and 6 should be the same clonotype

[Alexander230]

Hi, @ms3389!

I'm Aleksandr Popov, a developer of Immunarch package. Thank you for using our software!

I've made a branch 10x_filt_contigs_fix with a fix for this bug. Eventually it will be merged into dev, and then included into the next release of Immunarch.

You can install Immunarch version with this fix now with these commands:

install.packages(c("devtools", "pkgload"))
devtools::install_github("immunomind/immunarch", ref="10x_filt_contigs_fix")
devtools::reload(pkgload::inst("immunarch"))

Best regards, Aleksandr

[ms3389]

Thanks so much! Your package is fantastic!

[ms3389]

Hi Aleksandr,

I'm still seeing some issues that raise questions. For example if I load up this file into Immunarch: Patient 13 D7.csv

I get these clonotypes:

Line 1 and 10 is the same clonotype in the original 10x clonotype file marked as clonotype1 clonotype1_consensus_1 but loaded in Immunarch they are clearly marked as different clonotypes. There are other clonotypes with the same behavior in that same file.

Do you have any advice on how to approach this? Is this a bug?

Thank you.