Sanchit Misra

Parallel Mutual Information Based Construction of Genome-Scale Networks on the Intel®Xeon Phi™ Coprocessor

Construction of whole-genome networks from large-scale gene expression data is an important problem in systems biology. While several techniques have been developed, most cannot handle network reconstruction at the whole-genome scale, and the few that can, require large clusters. In this paper, we present a solution on the Intel Xeon Phi coprocessor, taking advantage of its multi-level parallelism including many x86-based cores, multiple threads per core, and vector processing units. We also present a solution on the Intel® Xeon® processor. Our solution is based on TINGe, a fast parallel network reconstruction technique that uses mutual information and permutation testing for assessing statistical significance. We demonstrate the first ever inference of a plant whole genome regulatory network on a single chip by constructing a 15,575 gene network of the plant Arabidopsis thaliana from 3,137 microarray experiments in only 22 minutes. In addition, our optimization for parallelizing mutual information computation on the Intel Xeon Phi coprocessor holds out lessons that are applicable to other domains.

Parallel Mutual Information Based Construction of Whole-Genome Networks on the Intel (R) Xeon Phi (TM) Coprocessor

Construction of whole-genome networks from large-scale gene expression data is an important problem in systems biology. While several techniques have been developed, most cannot handle network reconstruction at the whole-genome scale, and the few that can, require large clusters. In this paper, we present a solution on the Intel (R) Xeon Phi (TM) coprocessor, taking advantage of its multi-level parallelism including many x86-based cores, multiple threads per core, and vector processing units. We also present a solution on the Intel (R) Xeon (R) processor. Our solution is based on TINGe, a fast parallel network reconstruction technique that uses mutual information and permutation testing for assessing statistical significance. We demonstrate the first ever inference of a plant whole genome regulatory network on a single chip by constructing a 15,575 gene network of the plant Arabidopsis thaliana from 3,137 microarray experiments in only 22 minutes. In addition, our optimization for parallelizing mutual information computation on the Intel Xeon Phi coprocessor holds out lessons that are applicable to other domains.

muBLASTP: database-indexed protein sequence search on multicore CPUs

BackgroundThe Basic Local Alignment Search Tool (BLAST) is a fundamental program in the life sciences that searches databases for sequences that are most similar to a query sequence. Currently, the BLAST algorithm utilizes a query-indexed approach. Although many approaches suggest that sequence search with a database index can achieve much higher throughput (e.g., BLAT, SSAHA, and CAFE), they cannot deliver the same level of sensitivity as the query-indexed BLAST, i.e., NCBI BLAST, or they can only support nucleotide sequence search, e.g., MegaBLAST. Due to different challenges and characteristics between query indexing and database indexing, the existing techniques for query-indexed search cannot be used into database indexed search.ResultsmuBLASTP, a novel database-indexed BLAST for protein sequence search, delivers identical hits returned to NCBI BLAST. On Intel Haswell multicore CPUs, for a single query, the single-threaded muBLASTP achieves up to a 4.41-fold speedup for alignment stages, and up to a 1.75-fold end-to-end speedup over single-threaded NCBI BLAST. For a batch of queries, the multithreaded muBLASTP achieves up to a 5.7-fold speedups for alignment stages, and up to a 4.56-fold end-to-end speedup over multithreaded NCBI BLAST.ConclusionsWith a newly designed index structure for protein database and associated optimizations in BLASTP algorithm, we re-factored BLASTP algorithm for modern multicore processors that achieves much higher throughput with acceptable memory footprint for the database index.

Parallel bayesian network structure learning for genome-scale gene networks

Learning Bayesian networks is NP-hard. Even with recent progress in heuristic and parallel algorithms, modeling capabilities still fall short of the scale of the problems encountered. In this paper, we present a massively parallel method for Bayesian network structure learning, and demonstrate its capability by constructing genome-scale gene networks of the model plant Arabidopsis thaliana from over 168.5 million gene expression values. We report strong scaling efficiency of 75% and demonstrate scaling to 1.57 million cores of the Tianhe-2 supercomputer. Our results constitute three and five orders of magnitude increase over previously published results in the scale of data analyzed and computations performed, respectively. We achieve this through algorithmic innovations, using efficient techniques to distribute work across all compute nodes, all available processors and coprocessors on each node, all available threads on each processor and coprocessor, and vectorization techniques to maximize single thread performance.

Eliminating Irregularities of Protein Sequence Search on Multicore Architectures

Finding regions of local similarity between biological sequences is a fundamental task in computational biology. BLAST is the most widely-used tool for this purpose, but it suffers from irregularities due to its heuristic nature. To achieve fast search, recent approaches construct the index from the database instead of the input query. However, database indexing introduces more challenges in the design of index structure and algorithm, especially for data access through the memory hierarchy on modern multicore processors. In this paper, based on existing heuristic algorithms, we design and develop a database indexed BLAST with the identical sensitivity as query indexed BLAST (i.e., NCBI-BLAST). Then, we identify that existing heuristic algorithms of BLAST can result in serious irregularities in database indexed search. To eliminate irregularities in BLAST algorithm, we propose muBLASTP, that uses multiple optimizations to improve data locality and parallel efficiency for multicore architectures and multi-node systems. Experiments on a single node demonstrate up to a 5.1-fold speedup over the multi-threaded NCBI BLAST. For the inter-node parallelism, we achieve nearly linear scaling on up to 128 nodes and gain up to 8.9-fold speedup over mpiBLAST.

Scaling up Hartree-Fock calculations on Tianhe-2

This paper presents a new optimized and scalable code for Hartree–Fock self-consistent field iterations. Goals of the code design include scalability to large numbers of nodes, and the capability to simultaneously use CPUs and Intel Xeon Phi coprocessors. Issues we encountered as we optimized and scaled up the code on Tianhe-2 are described and addressed. A major issue is load balance, which is made challenging due to integral screening. We describe a general framework for finding a well-balanced static partitioning of the load in the presence of screening. Work stealing is used to polish the load balance. Performance results are shown on Stampede and Tianhe-2 supercomputers. Scalability is demonstrated on large simulations involving 2938 atoms and 27,394 basis functions, utilizing 8100 nodes of Tianhe-2.

Dtree: Dynamic Task Scheduling at Petascale

Irregular applications are challenging to scale on supercomputers due to the difficulty of balancing load across large numbers of nodes. This challenge is exacerbated by the increasing heterogeneity of modern supercomputers in which nodes often contain multiple processors and coprocessors operating at different speeds, and with differing core and thread counts. We present Dtree, a dynamic task scheduler designed to address this challenge. Dtree shows close to optimal results for a class of HPC applications, improving time-to-solution by achieving near-perfect load balance while consuming negligible resources. We demonstrate Dtree’s effectiveness on up to 77,824 heterogeneous cores of the TACC Stampede supercomputer with two different petascale HPC applications: ParaBLe, which performs large-scale Bayesian network structure learning, and GTFock, which implements Fock matrix construction, an essential and expensive step in quantum chemistry codes. For ParaBLe, we show improved performance while eliminating the complexity of managing heterogeneity. For GTFock, we match the most recently published performance without using any application-specific optimizations for data access patterns (such as the task distribution design for communication reduction) that enabled that performance. We also show that Dtree can distribute from tens of thousands to hundreds of millions of irregular tasks across up to 1024 nodes with minimal overhead, while balancing load to within 2 % of optimal.

Performance extraction and suitability analysis of multi- and many-core architectures for next generation sequencing secondary analysis

High-throughput next generation sequencers (NGS) can rapidly read billions of short DNA fragments, called reads, at low cost. Moreover, their throughput is increasing and cost is decreasing at rates much faster than the Moores law. This demands commensurate acceleration for NGS secondary analysis that process the reads to identify variations between genomes. Conventional architectural improvements can at best improve performance at the rate of Moores law even if the software tools efficiently utilize the underlying architecture. Unfortunately, most of the dozens of software products developed for this purpose fail to exploit the underlying architecture well. Therefore, to match the pace of development of the sequencers, we will need architecture that is more tailored for the computational requirements of NGS secondary analysis as well as software that uses the architecture optimally. To this end, in this work, we study the performance characteristics of NGS secondary analysis and investigate the suitability of modern Intel Xeon and Xeon Phi processors for the same. To keep the study manageable, we rely on recent studies that attribute a majority of the run-time to a few key kernels. We present detailed optimization efforts to accelerate these kernels on the latest Intel Xeon and Xeon Phi processors with the goal of extracting maximum performance. A comparison of our optimized implementations, along with published results on GPGPU implementations, shows that our optimized implementations on the Skylake processors yield highest performance. We also present an in-depth analysis of the key kernels and identify their performance characteristics and bottlenecks to inform future architectural designs.

Identification of Significant Computational Building Blocks through Comprehensive Deep Dive of NGS Secondary Analysis Methods

Rapid advancements in next generation sequencing technologies have greatly improved the throughput of sequencing and reduced the cost to under

Parallel Mutual Information Based Construction of Genome-Scale Networks on the Intel

1000 per genome propelling ambitious projects across the globe that are pursuing sequencing million or more genomes. In addition, the sequencing throughput is increasing and the cost is decreasing at a rate much faster than the Moore’s law. This necessitates equivalent rate of acceleration of NGS secondary analysis that assembles the reads into full genomes and identifies variants between genomes. Conventional improvement in hardware can at best help accelerate this according to the Moore’s law if the corresponding software is able to use the hardware efficiently. This is currently not the case for majority of the dozens of software tools used for NGS secondary analysis. Thus, to keep pace with the rate of advancement of sequencers, we need – 1) hardware that is designed taking into account the computational requirements of NGS secondary analysis and 2) software tools that use the hardware efficiently. In this work, we take the first step towards that goal by identifying the computational requirements of NGS secondary analysis. We surveyed dozens of software tools from all the three major problems in secondary analysis – sequence mapping, de novo assembly, and variant calling – to select seven popular tools and a workflow for an in depth analysis. We performed runtime profiling of the tools using multiple real datasets to find that the majority of the runtime is dominated by just four building blocks – Smith Waterman alignment, FM-index based sequence search, Debruijn graph construction and traversal and pairwise hidden markov model algorithm. Together, these building blocks cover 80.5%-98.2% of the runtime for sequence mapping, 63.9%-99.4% of the runtime for De novo assembly, and 72%-93% of the runtime for variant calling. The beauty of this result is that by just tailoring our software and hardware for these building blocks, we can get a major performance improvement of NGS secondary analysis.