Back

Accelerating Identification of Chromatin Accessibility from noisy ATAC-seq Data using Modern CPUs

Chaudhary, N.; Misra, S.; Kalamkar, D.; Heinecke, A.; Georganas, E.; Ziv, B.; Adelman, M.; Kaul, B.

2021-09-30 genomics
10.1101/2021.09.28.462099 bioRxiv
Show abstract

Identifying accessible chromatin regions is a fundamental problem in epigenomics with ATAC-seq being a commonly used assay. Exponential rise in single cell ATAC-seq experiments has made it critical to accelerate processing of ATAC-seq data. ATAC-seq data can have a low signal-to-noise ratio for various reasons including low coverage or low cell count. To denoise and identify accessible chromatin regions from noisy ATAC-seq data, use of deep learning on 1D data - using large filter sizes, long tensor widths, and/or dilation - has recently been proposed. Here, we present ways to accelerate the end-to-end training performance of these deep learning based methods using CPUs. We evaluate our approach on the recently released AtacWorks toolkit. Compared to an Nvidia DGX-1 box with 8 V100 GPUs, we get up to 2.27x speedup using just 16 CPU sockets. To achieve this, we build an efficient 1D dilated convolution layer and demonstrate reduced precision (BFloat16) training.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

1
Genome Research
409 papers in training set
Top 0.1%
27.8%
2
Nature Communications
4913 papers in training set
Top 9%
14.8%
3
Bioinformatics
1061 papers in training set
Top 2%
12.4%
50% of probability mass above
4
Nature Methods
336 papers in training set
Top 2%
6.4%
5
Genome Biology
555 papers in training set
Top 1%
6.4%
6
Nature Biotechnology
147 papers in training set
Top 2%
4.0%
7
Nucleic Acids Research
1128 papers in training set
Top 7%
3.1%
8
NAR Genomics and Bioinformatics
214 papers in training set
Top 1%
2.6%
9
PLOS Computational Biology
1633 papers in training set
Top 17%
1.7%
10
BMC Bioinformatics
383 papers in training set
Top 5%
1.7%
11
Communications Biology
886 papers in training set
Top 11%
1.5%
12
Cell Reports Methods
141 papers in training set
Top 3%
1.5%
13
GigaScience
172 papers in training set
Top 2%
1.0%
14
BMC Genomics
328 papers in training set
Top 4%
1.0%
15
Nature Computational Science
50 papers in training set
Top 1%
0.8%
16
Cell Genomics
162 papers in training set
Top 6%
0.7%
17
PLOS Genetics
756 papers in training set
Top 15%
0.7%
18
Frontiers in Genetics
197 papers in training set
Top 10%
0.7%
19
G3: Genes, Genomes, Genetics
222 papers in training set
Top 1.0%
0.7%
20
Scientific Reports
3102 papers in training set
Top 75%
0.7%
21
iScience
1063 papers in training set
Top 32%
0.7%
22
Nature Genetics
240 papers in training set
Top 8%
0.7%
23
Computational and Structural Biotechnology Journal
216 papers in training set
Top 11%
0.6%
24
The American Journal of Human Genetics
206 papers in training set
Top 5%
0.5%