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

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.

One‐dimensional atrous conv‐net based architecture for automatic diagnosis of epilepsy using electroencephalography signals and its brain–computer interface applications

“…Table 4 shows the achieved results for cases 1, 2, and 4 results for 150 epochs. Case 3 results are not reported due to the insufficient input shape of (Chaudhary et al, 2021; Osalusi et al, 2018) received at conv1d 7th filter of the model. It produced an output shape with a zero or negative value which was not compatible with the next filter.…”

“…Second, it increases the span of the filter without increasing the number of weight parameters in them (Chaudhary et al, 2021). It helps in maintaining the order of the data which is vital for sequential data analysis.…”

“…Max pooling is considered beneficial for spatial transforms and hence has been utilized in the proposed architecture (Chaudhary et al, 2021). Similarly, the batch normalization layer is known to regularize a DL architecture.…”