seq2HLA computationally determines human leukocyte antigen (HLA) genotypes of a sample. It takes standard RNA-Seq sequence reads in fastq format as input, uses a bowtie index comprising all HLA alleles and outputs the most likely HLA class I and class II genotypes (in 4 digit resolution), a p-value for each call, and the expression of each class.

References:

• Boegel, Sebastian, Martin Löwer, Michael Schäfer, Thomas Bukur, Jos De Graaf, Valesca Boisguérin, Özlem Türeci, Mustafa Diken, John C. Castle, and Ugur Sahin. HLA typing from RNA-Seq sequence reads. Genome medicine 4 (2013): 1-12.

• seq2HLA Main Site

• Module Name: seq2hla (see the modules page for more information)
• Multithreaded: use -p
• Environment variables set
  • SEQ2HLA_HOME
• Example files in $SEQ2HLA_HOME/test

Allocate an interactive session and run the program.
Sample session (user input in bold):

[user@biowulf]$ sinteractive --cpus-per-task 8 --mem 32g --gres lscratch:10
salloc.exe: Pending job allocation 46116226
salloc.exe: job 46116226 queued and waiting for resources
salloc.exe: job 46116226 has been allocated resources
salloc.exe: Granted job allocation 46116226
salloc.exe: Waiting for resource configuration
salloc.exe: Nodes cn3144 are ready for job
[user@cn3144 ~]$ module load seq2hla
[user@cn3144 ~]$ cd /lscratch/$SLURM_JOB_ID
[user@cn3144 46116226]$ cp $SEQ2HLA_HOME/test/*.fastq.gz .
[user@cn3144 46116226]$ seq2HLA -1 SRR4300096_1.fastq.gz -2 SRR4300096_2.fastq.gz -r test -p $SLURM_CPUS_PER_TASK
Now running seq2HLA version 2.3!
Input is a gipped file .....
The read length of your input fastq was determined to be 100, so 2 mismatches will be allowed and 8 threads will be used by bowtie.
----------HLA class I------------
>---classical HLA alleles---
First iteration starts....
Mapping ......
# reads processed: 34632623
# reads with at least one reported alignment: 10689 (0.03%)
# reads that failed to align: 34621934 (99.97%)
Reported 853322 paired-end alignments to 1 output stream(s)

Calculation of first digital haplotype.....
The digital haplotype is written into test-ClassI-class.digitalhaplotype1
1st iteration done.
Now removing reads that mapped to the three top-scoring groups .......
Second iterations starts .....
 Mapping ......
# reads processed: 4332
# reads with at least one reported alignment: 4332 (100.00%)
# reads that failed to align: 0 (0.00%)
Reported 363543 paired-end alignments to 1 output stream(s)
Calculation of second digital haplotype.....
The digital haplotype is written into test-ClassI-class.digitalhaplotype2
2nd iteration done.
-----------2 digit typing results-------------
#Locus	Allele 1	Confidence	Allele 2	Confidence
A	A*25	0.0006404339	A*01	0.01656387
B	B*08	0.0008702224	B*18	1.98204e-05
C	C*12	7.154939e-05	C*07	0.006626664
Calculation of locus-specific expression ...
test-ClassI-class.bowtielog
A: 76.22 RPKM
C: 91.62 RPKM
B: 96.65 RPKM
-----------4 digit typing results-------------
#Locus	Allele 1	Confidence	Allele 2	Confidence
A	A*25:01	0.0006404339	A*01:01	0.0179885
B	B*08:01	0.0008702224	B*18:01	1.98204e-05
C	C*12:03	7.51191e-05	C*07:01'	0.006944861
----------HLA class I------------
>---nonclassical HLA alleles---
First iteration starts....
Mapping ......
# reads processed: 34632623
# reads with at least one reported alignment: 4465 (0.01%)
# reads that failed to align: 34628158 (99.99%)
Reported 35369 paired-end alignments to 1 output stream(s)

Calculation of first digital haplotype.....
The digital haplotype is written into test-ClassI-nonclass.digitalhaplotype1
1st iteration done.
Now removing reads that mapped to the three top-scoring groups .......
Second iterations starts .....
 Mapping ......
Warning: Could not find any reads in "test-ClassI-nonclass-2nditeration_1.fq"
Warning: Could not find any reads in "test-ClassI-nonclass-2nditeration_2.fq"
# reads processed: 0
# reads with at least one reported alignment: 0 (0.00%)
# reads that failed to align: 0 (0.00%)
No alignments
Calculation of second digital haplotype.....
The digital haplotype is written into test-ClassI-nonclass.digitalhaplotype2
2nd iteration done.
-----------2 digit typing results-------------
#Locus	Allele 1	Confidence	Allele 2	Confidence
E	E*01	NA	hoz("E*01")	NA
F	F*01	NA	hoz("F*01")	NA
G	no	NA	hoz("G*01")	NA
H	H*02	0	hoz("H*02")	NA
J	J*01	NA	hoz("J*01")	NA
K	K*01	NA	hoz("K*01")	NA
L	L*01	NA	hoz("L*01")	NA
P	no	NA	hoz("P*02")	NA
V	no	NA	hoz("V*01")	NA
Calculation of locus-specific expression ...
test-ClassI-nonclass.bowtielog
E: 115.75 RPKM
G: 0.0 RPKM
F: 18.44 RPKM
H: 25.7 RPKM
K: 0.16 RPKM
J: 0.03 RPKM
L: 0.11 RPKM
P: 0.0 RPKM
V: 0.0 RPKM
# reads processed: 4465
# reads with at least one reported alignment: 4465 (100.00%)
# reads that failed to align: 0 (0.00%)
Reported 35369 paired-end alignments to 1 output stream(s)
The digital haplotype is written into test-ClassI-nonclass.digitalhaplotype3
-----------4 digit typing results-------------
#Locus	Allele 1	Confidence	Allele 2	Confidence
E	E*01:01	NA	E*01:01	NA
F	F*01:01'	NA	F*01:01	NA
G	no	NA	no	NA
H	H*02:06	0.0	H*02:06	NA
J	J*01:01	NA	J*01:01	NA
K	K*01:01	NA	K*01:01	NA
L	L*01:01	NA	L*01:01	NA
P	no	NA	no	NA
V	no	NA	no	NA
----------HLA class II------------
First iteration starts....
Mapping ......
# reads processed: 34632623
# reads with at least one reported alignment: 1889 (0.01%)
# reads that failed to align: 34630734 (99.99%)
Reported 57508 paired-end alignments to 1 output stream(s)

Calculation of first digital haplotype.....
The digital haplotype is written into test-ClassII.digitalhaplotype1
1st iteration done.
Now removing reads that mapped to the three top-scoring groups .......
Second iterations starts .....
 Mapping ......
# reads processed: 395
# reads with at least one reported alignment: 395 (100.00%)
# reads that failed to align: 0 (0.00%)
Reported 19773 paired-end alignments to 1 output stream(s)
Calculation of second digital haplotype.....
The digital haplotype is written into test-ClassII.digitalhaplotype2
2nd iteration done.
-----------2 digit typing results-------------
#Locus	Allele 1	Confidence	Allele 2	Confidence
DQA1	DQA1*03	1.534371e-05	DQA1*05	0.2468488
DQB1	DQB1*02	0.2004472	DQB1*03	0.3331947
DRB1	DRB1*04	0.0006670405	DRB1*03	0.04627267
DRA	DRA*01	NA	hoz("DRA*01")	NA
DPA1	DPA1*01	0.3854021	hoz("DPA1*02")	0.0574801
DPB1	DPB1*105	1.045282e-07	DPB1*04	0.1606171
Calculation of locus-specific expression ...
test-ClassII.bowtielog
DQB1: 2.54 RPKM
DQA1: 6.5 RPKM
DRB1: 26.2 RPKM
DPB1: 12.76 RPKM
DRA: 40.35 RPKM
DPA1: 17.21 RPKM
# reads processed: 1889
# reads with at least one reported alignment: 1889 (100.00%)
# reads that failed to align: 0 (0.00%)
Reported 57508 paired-end alignments to 1 output stream(s)
The digital haplotype is written into test-ClassII.digitalhaplotype3
-----------4 digit typing results-------------
#Locus	Allele 1	Confidence	Allele 2	Confidence
DQA1	DQA1*03:02'	1.534371e-05	DQA1*05:01	0.2468488
DQB1	DQB1*02:01'	0.2004472	DQB1*03:03'	0.3331947
DRB1	DRB1*04:01'	0.0006670405	DRB1*03:01	0.04627267
DRA	DRA*01:01	NA	DRA*01:01	NA
DPA1	DPA1*01:03	0.3854021	DPA1*01:03	0.0574801
DPB1	DPB1*105:01	1.045282e-07	DPB1*04:01	0.1606171

[user@cn3144 46116226]$ exit
salloc.exe: Relinquishing job allocation 46116226
[user@biowulf ~]$

Create a batch input file (e.g. seq2hla.sh). For example:

#!/bin/bash
set -e
module load seq2hla
seq2HLA -1 SRR4300096_1.fastq.gz -2 SRR4300096_2.fastq.gz -r test -p $SLURM_CPUS_PER_TASK

Submit this job using the Slurm sbatch command.

sbatch --cpus-per-task=# [--mem=#] seq2hla.sh

Create a swarmfile (e.g. seq2hla.swarm). For example:

seq2HLA -1 sample1_1.fastq.gz -2 sample1_2.fastq.gz -r sample1 -p $SLURM_CPUS_PER_TASK
seq2HLA -1 sample2_1.fastq.gz -2 sample2_2.fastq.gz -r sample2 -p $SLURM_CPUS_PER_TASK
seq2HLA -1 sample3_1.fastq.gz -2 sample3_2.fastq.gz -r sample3 -p $SLURM_CPUS_PER_TASK
seq2HLA -1 sample4_1.fastq.gz -2 sample4_2.fastq.gz -r sample4 -p $SLURM_CPUS_PER_TASK

Submit this job using the swarm command.

swarm -f seq2hla.swarm [-g #] -t # --module seq2hla