Ananth Kalyanaraman

Network-on-Chip Hardware Accelerators for Biological Sequence Alignment

The most pervasive compute operation carried out in almost all bioinformatics applications is pairwise sequence homology detection (or sequence alignment). Due to exponentially growing sequence databases, computing this operation at a large-scale is becoming expensive. An effective approach to speed up this operation is to integrate a very high number of processing elements in a single chip so that the massive scales of fine-grain parallelism inherent in several bioinformatics applications can be exploited efficiently. Network-on-chip (NoC) is a very efficient method to achieve such large-scale integration. In this work, we propose to bridge the gap between data generation and processing in bioinformatics applications by designing NoC architectures for the sequence alignment operation. Specifically, we 1) propose optimized NoC architectures for different sequence alignment algorithms that were originally designed for distributed memory parallel computers and 2) provide a thorough comparative evaluation of their respective performance and energy dissipation. While accelerators using other hardware architectures such as FPGA, general purpose graphics processing unit (GPU), and the cell broadband engine (CBE) have been previously designed for sequence alignment, the NoC paradigm enables integration of a much larger number of processing elements on a single chip and also offers a higher degree of flexibility in placing them along the die to suit the underlying algorithm. The results show that our NoC-based implementations can provide above 102-103-fold speedup over other hardware accelerators and above 104-fold speedup over traditional CPU architectures. This is significant because it will drastically reduce the time required to perform the millions of alignment operations that are typical in large-scale bioinformatics projects. To the best of our knowledge, this work embodies the first attempt to accelerate a bioinformatics application using NoC.

Detailed analysis of a contiguous 22-Mb region of the maize genome.

Most of our understanding of plant genome structure and evolution has come from the careful annotation of small (e.g., 100 kb) sequenced genomic regions or from automated annotation of complete genome sequences. Here, we sequenced and carefully annotated a contiguous 22 Mb region of maize chromosome 4 using an improved pseudomolecule for annotation. The sequence segment was comprehensively ordered, oriented, and confirmed using the maize optical map. Nearly 84% of the sequence is composed of transposable elements (TEs) that are mostly nested within each other, of which most families are low-copy. We identified 544 gene models using multiple levels of evidence, as well as five miRNA genes. Gene fragments, many captured by TEs, are prevalent within this region. Elimination of gene redundancy from a tetraploid maize ancestor that originated a few million years ago is responsible in this region for most disruptions of synteny with sorghum and rice. Consistent with other sub-genomic analyses in maize, small RNA mapping showed that many small RNAs match TEs and that most TEs match small RNAs. These results, performed on ∼1% of the maize genome, demonstrate the feasibility of refining the B73 RefGen_v1 genome assembly by incorporating optical map, high-resolution genetic map, and comparative genomic data sets. Such improvements, along with those of gene and repeat annotation, will serve to promote future functional genomic and phylogenomic research in maize and other grasses.

Hardware accelerators for biocomputing: A survey

Computing research has become a vital cog in the machinery required to drive biological discovery. Computing has made possible significant achievements over the last decade, especially in the genomics sector. An emerging area is the investigation of hardware accelerators for speeding up the massive scale of computation needed in large-scale biocomputing applications. Various hardware platforms, such as FPGA, Graphics Processing Unit (GPU), the Cell Broadband Engine (CBE) and multi-core processors are being explored. In this paper, we present a survey of hardware accelerators for biocomputing by choosing a representative set of each.

BioEarth: Envisioning and developing a new regional earth system model to inform natural and agricultural resource management

As managers of agricultural and natural resources are confronted with uncertainties in global change impacts, the complexities associated with the interconnected cycling of nitrogen, carbon, and water present daunting management challenges. Existing models provide detailed information on specific sub-systems (e.g., land, air, water, and economics). An increasing awareness of the unintended consequences of management decisions resulting from interconnectedness of these sub-systems, however, necessitates coupled regional earth system models (EaSMs). Decision makers’ needs and priorities can be integrated into the model design and development processes to enhance decision-making relevance and “usability” of EaSMs. BioEarth is a research initiative currently under development with a focus on the U.S. Pacific Northwest region that explores the coupling of multiple stand-alone EaSMs to generate usable information for resource decision-making. Direct engagement between model developers and non-academic stakeholders involved in resource and environmental management decisions throughout the model development process is a critical component of this effort. BioEarth utilizes a bottom-up approach for its land surface model that preserves fine spatial-scale sensitivities and lateral hydrologic connectivity, which makes it unique among many regional EaSMs. This paper describes the BioEarth initiative and highlights opportunities and challenges associated with coupling multiple stand-alone models to generate usable information for agricultural and natural resource decision-making.

Attenuation of virulence in an apicomplexan hemoparasite results in reduced genome diversity at the population level

BackgroundVirulence acquisition and loss is a dynamic adaptation of pathogens to thrive in changing milieus. We investigated the mechanisms of virulence loss at the whole genome level using Babesia bovis as a model apicomplexan in which genetically related attenuated parasites can be reliably derived from virulent parental strains in the natural host. We expected virulence loss to be accompanied by consistent changes at the gene level, and that such changes would be shared among attenuated parasites of diverse geographic and genetic background.ResultsSurprisingly, while single nucleotide polymorphisms in 14 genes distinguished all attenuated parasites from their virulent parental strains, all non-synonymous changes resulted in no deleterious amino acid modification that could consistently be associated with attenuation (or virulence) in this hemoparasite. Interestingly, however, attenuation significantly reduced the overall populations genome diversity with 81% of base pairs shared among attenuated strains, compared to only 60% of base pairs common among virulent parental parasites. There were significantly fewer genes that were unique to their geographical origins among the attenuated parasites, resulting in a simplified population structure among the attenuated strains.ConclusionsThis simplified structure includes reduced diversity of the variant erythrocyte surface 1 (ves) multigene family repertoire among attenuated parasites when compared to virulent parental strains, possibly suggesting that overall variance in large protein families such as Variant Erythrocyte Surface Antigens has a critical role in expression of the virulence phenotype. In addition, the results suggest that virulence (or attenuation) mechanisms may not be shared among all populations of parasites at the gene level, but instead may reflect expansion or contraction of the population structure in response to shifting milieus.

An efficient parallel approach for identifying protein families in large-scale metagenomic data sets

Metagenomics is the study of environmental microbial communities using state-of-the-art genomic tools. Recent advancements in high-throughput technologies have enabled the accumulation of large volumes of metagenomic data that was until a couple of years back was deemed impractical for generation. A primary bottleneck, however, is in the lack of scalable algorithms and open source software for large-scale data processing. In this paper, we present the design and implementation of a novel parallel approach to identify protein families from large-scale metagenomic data. Given a set of peptide sequences we reduce the problem to one of detecting arbitrarily-sized dense subgraphs from bipartite graphs. Our approach efficiently parallelizes this task on a distributed memory machine through a combination of divide-and-conquer and combinatorial pattern matching heuristic techniques. We present performance and quality results of extensively testing our implementation on 160 K randomly sampled sequences from the CAMERA environmental sequence database using 512 nodes of a BlueGene/L supercomputer.

Parallel Heuristics for Scalable Community Detection

Community detection has become a fundamental operation in numerous graph-theoretic applications. It is used to reveal natural divisions that exist within real world networks without imposing prior size or cardinality constraints on the set of communities. Despite its potential for application, there is only limited support for community detection on large-scale parallel computers, largely owing to the irregular and inherently sequential nature of the underlying heuristics. In this paper, we present parallelization heuristics for fast community detection using the Louvain method as the serial template. The Louvain method is an iterative heuristic for modularity optimization. Originally developed by Blondel et al. in 2008, the method has become increasingly popular owing to its ability to detect high modularity community partitions in a fast and memory-efficient manner. However, the method is also inherently sequential, thereby limiting its scalability. Here, we observe certain key properties of this method that present challenges for its parallelization, and consequently propose heuristics that are designed to break the sequential barrier. For evaluation purposes, we implemented our heuristics using OpenMP multithreading, and tested them over real world graphs derived from multiple application domains (e.g., internet, citation, biological). Compared to the serial Louvain implementation, our parallel implementation is able to produce community outputs with a higher modularity for most of the inputs tested, in comparable number of iterations, while providing real speedups of up to 8× using 32 threads. In addition, our parallel implementation was able to exhibit weak scaling properties on up to 32 threads.

Rosaceaous Genome Sequencing: Perspectives and Progress

The long-term goal of plant genomics is to identify, isolate and determine the function of plant genes that are associated with both vegetative and reproductive phenotypes. Most phenotypes require the coordinated activity and regulatory control of suites of genes over time and in precise positions within the plant. Until recently, the idea of establishing a comprehensive approach to isolate and characterize all the genes involved in any complex phenotype was a daunting one, and oftentimes it has been necessary to perform whole genome sequencing to obtain all of the gene sequences. The sequence of several plant genomes have been generated including the Arabidopsis, poplar and rice genomes, with the model legume Medicago and sorghum well underway. In addition, large amounts of expressed sequence tag (EST) information are being obtained for many other plants, including rosaceaous plants. The advances in genomics, informatics, and phylogenetics have been developed and refined by these reference projects, to the point that it is now thought that, in many cases, the vast majority of genes of a plant can be identified without the complete genome sequence, however the EST approach to gene identification does not provide valuable information regarding promoters and other non-coding regulatory elements. One of the first eukaryotic genomes to be completely sequenced was that of the small mustard species Arabidopsis thaliana. During the past decade, Arabidopsis has emerged as one of the most widely used model organisms for studying the biology of higher plants. Its genome was chosen for sequencing because it is highly compact, about 125 Mb, with little interspersed repetitive DNA (Arabidopsis Genome Initiative, 2000). However, since Arabidopsis is rather distantly related to the cereal crops that provide the bulk of the world food supply, the genome of rice has also been sequenced. Rice was chosen because, in addition to its importance as a food source for about one quarter of the human population, it has one of

Fast SVD Computations for Synchrophasor Algorithms

Many singular value decomposition (SVD) problems in power system computations require only a few largest singular values of a large-scale matrix for the analysis. This letter introduces two fast SVD approaches recently developed in other domains to power systems for speeding up phasor measurement unit (PMU) based online applications. The first method is a randomized SVD algorithm that accelerates computation by introducing a low-rank approximation of a given matrix through randomness. The second method is the augmented Lanczos bidiagonalization, an iterative Krylov subspace technique that computes sequences of projections of a given matrix onto low-dimensional subspaces. Both approaches are illustrated on SVD evaluation within an ambient oscillation monitoring algorithm, namely stochastic subspace identification (SSI).

Balanced Coloring for Parallel Computing Applications

Graph colouring is used to identify subsets of independent tasks in parallel scientific computing applications. Traditional colouring heuristics aim to reduce the number of colours used as that number also corresponds to the number of parallel steps in the application. However, if the color classes produced have a skew in their sizes, utilization of hardware resources becomes inefficient, especially for the smaller color classes. Equitable colouring is a theoretical formulation of colouring that guarantees a perfect balance among color classes, and its practical relaxation is referred to as balanced colouring. In this paper, we revisit the problem of balanced colouring in the context of parallel computing. The goal is to achieve a balanced colouring of an input graph without increasing the number of colours that an algorithm oblivious to balance would have used. We propose and study multiple heuristics that aim to achieve such a balanced colouring, present parallelization approaches for multi-core and manicure architectures, and cross-evaluate their effectiveness with respect to the quality of balance achieved and performance. Furthermore, we study the impact of the proposed balanced colouring heuristics on a concrete application - viz. parallel community detection, which is an example of an irregular application. The thorough treatment of balanced colouring presented in this paper from algorithms to application is expected to serve as a valuable resource to parallel application developers who seek to improve parallel performance of their applications using colouring.