R package – a GPU-based extension for the HIBAG package
-
OpenCL C library and header files (https://www.khronos.org)
-
HIBAG R package (≥ v1.28.0)
Changes in v0.99.0 (Y2021):
-
new implementation using half and mixed precisions
-
a new function
hlaAttrBagging_MultiGPU()to leverage multiple GPU devices
v0.9.0 (Y2018):
- initial GPU-based implementation for the HIBAG algorithm
| CPU / GPU | Precision | Factor |
|---|---|---|
| CPU (AVX2, 1 thread) | double | 1 |
| CPU (AVX2, 20 threads) | double | 14.7 |
| 1x NVIDIA Tesla T4 | half | 71.8 |
| 1x NVIDIA Tesla V100 | half | 74.7 |
| 1x NVIDIA Tesla T4 | mixed | 61.2 |
| 1x NVIDIA Tesla V100 | mixed | 68.3 |
| 1x NVIDIA Tesla T4 | single | 52.4 |
| 1x NVIDIA Tesla V100 | single | 61.3 |
| 1x NVIDIA Tesla V100 | double | 22.7 |
| CPU / GPU | Precision | Factor |
|---|---|---|
| CPU (AVX2, 1 thread) | double | 1 |
| CPU (AVX2, 20 threads) | double | 17.5 |
| 1x NVIDIA Tesla T4 | half | 101.6 |
| 1x NVIDIA Tesla V100 | half | 125.2 |
| 1x NVIDIA Tesla T4 | mixed | 72.0 |
| 1x NVIDIA Tesla V100 | mixed | 98.9 |
| 1x NVIDIA Tesla T4 | single | 60.8 |
| 1x NVIDIA Tesla V100 | single | 92.4 |
| 1x NVIDIA Tesla V100 | double | 18.4 |
† ‘mixed’ is a mixed precision between half and single
† models built on HLA-A, -B, -C, -DRB1 using HIBAG v1.26.1 and HIBAG.gpu v0.99.0, and the average is reported
† CPU (AVX2, 1/20 threads), optimization with Intel AVX2 instruction, using Intel(R) Xeon(R) Gold 6248 2.50GHz, 20 cores (Cascade Lake)
† This work was made possible, in part, through HPC time donated by Microway, Inc. We gratefully acknowledge Microway for providing access to their GPU-accelerated compute cluster (http://www.microway.com/gpu-test-drive/).
-
Zheng, X. et al. HIBAG-HLA genotype imputation with attribute bagging. Pharmacogenomics Journal 14, 192-200 (2014). http://dx.doi.org/10.1038/tpj.2013.18
-
Zheng, X. (2018) Imputation-Based HLA Typing with SNPs in GWAS Studies. In: Boegel S. (eds) HLA Typing. Methods in Molecular Biology, Vol 1802. Humana Press, New York, NY. https://doi.org/10.1007/978-1-4939-8546-3_11
- Development version from Github:
library("devtools")
install_github("zhengxwen/HIBAG.gpu")The install_github() approach requires that you build from source, i.e. make and compilers must be installed on your system -- see the R FAQ for your operating system; you may also need to install dependencies manually.
- Install the package from the source code: download the source code
wget --no-check-certificate https://github.com/zhengxwen/HIBAG.gpu/tarball/master -O HIBAG.gpu_latest.tar.gz
## or ##
curl -L https://github.com/zhengxwen/HIBAG.gpu/tarball/master/ -o HIBAG.gpu_latest.tar.gz
## Install ##
R CMD INSTALL HIBAG.gpu_latest.tar.gzlibrary(HIBAG.gpu)
## Loading required package: HIBAG
## HIBAG (HLA Genotype Imputation with Attribute Bagging)
## Kernel Version: v1.5 (64-bit, AVX2)
##
## Available OpenCL device(s):
## Dev #1: AMD, AMD Radeon Pro 560X Compute Engine, OpenCL 1.2
##
## Using Device #1: AMD, AMD Radeon Pro 560X Compute Engine
## Device Version: OpenCL 1.2 (>= v1.2: YES)
## Driver Version: 1.2 (Jan 12 2021 22:17:03)
## EXTENSION cl_khr_global_int32_base_atomics: YES
## EXTENSION cl_khr_fp64: YES
## EXTENSION cl_khr_int64_base_atomics: NO
## EXTENSION cl_khr_global_int64_base_atomics: NO
## CL_DEVICE_GLOBAL_MEM_SIZE: 4,294,967,296
## CL_DEVICE_MAX_MEM_ALLOC_SIZE: 1,073,741,824
## CL_DEVICE_MAX_COMPUTE_UNITS: 16
## CL_DEVICE_MAX_WORK_GROUP_SIZE: 256
## CL_DEVICE_MAX_WORK_ITEM_DIMENSIONS: 3
## CL_DEVICE_MAX_WORK_ITEM_SIZES: 256,256,256
## CL_DEVICE_LOCAL_MEM_SIZE: 32768
## CL_DEVICE_ADDRESS_BITS: 32
## CL_KERNEL_WORK_GROUP_SIZE: 256
## CL_KERNEL_PREFERRED_WORK_GROUP_SIZE_MULTIPLE: 64
## local work size: 256 (Dim1), 16x16 (Dim2)
## atom_cmpxchg (enable cl_khr_int64_base_atomics): OK
## GPU device supports double-precision floating-point numbers
## Training uses a mixed precision between half and float in GPU
## By default, prediction uses 32-bit floating-point numbers in GPU (since EXTENSION cl_khr_int64_base_atomics: NO).# make a "hlaAlleleClass" object
hla.id <- "A"
hla <- hlaAllele(HLA_Type_Table$sample.id,
H1 = HLA_Type_Table[, paste(hla.id, ".1", sep="")],
H2 = HLA_Type_Table[, paste(hla.id, ".2", sep="")],
locus=hla.id, assembly="hg19")
# SNP predictors within the flanking region on each side
region <- 500 # kb
snpid <- hlaFlankingSNP(HapMap_CEU_Geno$snp.id, HapMap_CEU_Geno$snp.position,
hla.id, region*1000, assembly="hg19")
length(snpid)
# training and validation genotypes
train.geno <- hlaGenoSubset(HapMap_CEU_Geno, snp.sel=match(snpid, HapMap_CEU_Geno$snp.id))
summary(train.geno)
# train a HIBAG model
set.seed(100)
model <- hlaAttrBagging_gpu(hla, train.geno, nclassifier=100)
summary(model)## Gene: A
## Training dataset: 60 samples X 266 SNPs
## # of HLA alleles: 14
## # of individual classifiers: 100
## total # of SNPs used: 239
## avg. # of SNPs in an individual classifier: 15.60
## (sd: 2.73, min: 10, max: 24, median: 16.00)
## avg. # of haplotypes in an individual classifier: 40.44
## (sd: 17.46, min: 18, max: 105, median: 36.00)
## avg. out-of-bag accuracy: 93.00%
## (sd: 4.94%, min: 78.95%, max: 100.00%, median: 93.48%)
## Matching proportion:
## Min. 0.1% Qu. 1% Qu. 1st Qu. Median 3rd Qu.
## 0.0004657777 0.0004773751 0.0005817518 0.0040887403 0.0112166087 0.0282807795
## Max. Mean SD
## 0.4263384035 0.0393261556 0.0919757944
## Genome assembly: hg19
# validation
pred <- hlaPredict_gpu(model, train.geno)
summary(pred)
# compare
(comp <- hlaCompareAllele(hla, pred, call.threshold=0)$overall)## total.num.ind crt.num.ind crt.num.haplo acc.ind acc.haplo call.threshold
## 60 59 119 0.9833333 0.9916667 0
## n.call call.rate
## 60 1