Maksudul Alam | Virginia Tech (original) (raw)

Papers by Maksudul Alam

Data Science and Engineering, 2019

A novel parallel algorithm is presented for generating random scale-free networks using the prefe... more A novel parallel algorithm is presented for generating random scale-free networks using the preferential attachment model. The algorithm, named cuPPA, is custom-designed for "single instruction multiple data (SIMD)" style of parallel processing supported by modern processors such as graphical processing units (GPUs). To the best of our knowledge, our algorithm is the first to exploit GPUs, and also the fastest implementation available today, to generate scale-free networks using the preferential attachment model. A detailed performance study is presented to understand the scalability and runtime characteristics of the cuPPA algorithm. Also another version of the algorithm called cuPPA-Hash tailored for multiple GPUs is presented. On a single GPU, the original cuPPA algorithm delivers the best performance, but is challenging to port to multi-GPU implementation. For multi-GPU implementation, cuPPA-Hash has been used as the parallel algorithm to achieve a perfect linear speedup up to 4 GPUs. In one of the best cases, when executed on an NVidia GeForce 1080 GPU, the original cuPPA generates a scale-free network of two billion edges in less than 3 s. On multi-GPU platforms, cuPPA-Hash generates a scale-free network of 16 billion edges in less than 7 s using a machine consisting of 4 NVidia Tesla P100 GPUs.

Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis, 2013

Recently, there has been substantial interest in the study of various random networks as mathemat... more Recently, there has been substantial interest in the study of various random networks as mathematical models of complex systems. As these complex systems grow larger, the ability to generate progressively large random networks becomes all the more important. This motivates the need for efficient parallel algorithms for generating such networks. Naive parallelization of the sequential algorithms for generating random networks may not work due to the dependencies among the edges and the possibility of creating duplicate (parallel) edges. In this paper, we present MPI-based distributed memory parallel algorithms for generating random scale-free networks using the preferential-attachment model. Our algorithms scale very well to a large number of processors and provide almost linear speedups. The algorithms can generate scale-free networks with 50 billion edges in 123 seconds using 768 processors.

PLoS ONE, 2013

T helper (Th) cells play a major role in the immune response and pathology at the gastric mucosa ... more T helper (Th) cells play a major role in the immune response and pathology at the gastric mucosa during Helicobacter pylori infection. There is a limited mechanistic understanding regarding the contributions of CD4+ T cell subsets to gastritis development during H. pylori colonization. We used two computational approaches: ordinary differential equation (ODE)based and agent-based modeling (ABM) to study the mechanisms underlying cellular immune responses to H. pylori and how CD4+ T cell subsets influenced initiation, progression and outcome of disease. To calibrate the model, in vivo experimentation was performed by infecting C57BL/6 mice intragastrically with H. pylori and assaying immune cell subsets in the stomach and gastric lymph nodes (GLN) on days 0, 7, 14, 30 and 60 post-infection. Our computational model reproduced the dynamics of effector and regulatory pathways in the gastric lamina propria (LP) in silico. Simulation results show the induction of a Th17 response and a dominant Th1 response, together with a regulatory response characterized by high levels of mucosal Treg) cells. We also investigated the potential role of peroxisome proliferator-activated receptor c (PPARc) activation on the modulation of host responses to H. pylori by using loss-of-function approaches. Specifically, in silico results showed a predominance of Th1 and Th17 cells in the stomach of the cell-specific PPARc knockout system when compared to the wild-type simulation. Spatio-temporal, object-oriented ABM approaches suggested similar dynamics in induction of host responses showing analogous T cell distributions to ODE modeling and facilitated tracking lesion formation. In addition, sensitivity analysis predicted a crucial contribution of Th1 and Th17 effector responses as mediators of histopathological changes in the gastric mucosa during chronic stages of infection, which were experimentally validated in mice. These integrated immunoinformatics approaches characterized the induction of mucosal effector and regulatory pathways controlled by PPARc during H. pylori infection affecting disease outcomes.

2012 IEEE 8th International Conference on E-Science, 2012

Networks are an effective abstraction for representing real systems. Consequently, network scienc... more Networks are an effective abstraction for representing real systems. Consequently, network science is increasingly used in academia and industry to solve problems in many fields. Computations that determine structure properties and dynamical behaviors of networks are useful because they give insights into the characteristics of real systems. We introduce a newly built and deployed cyberinfrastructure for network science (CINET) that performs such computations, with the following features: (i) it offers realistic networks from the literature and various random and deterministic network generators; (ii) it provides many algorithmic modules and measures to study and characterize networks; (iii) it is designed for efficient execution of complex algorithms on distributed high performance computers so that they scale to large networks; and (iv) it is hosted with web interfaces so that those without direct access to high performance computing resources and those who are not computing experts can still reap the system benefits. It is a combination of application design and cyberinfrastructure that makes these features possible. To our knowledge, these capabilities collectively make CINET novel. We describe the system and illustrative use cases, with a focus on the CINET user.

2008 International Conference on BioMedical Engineering and Informatics, 2008

Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Construct... more Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Constructing a pair of haplotypes from aligned and overlapping but intermixed and erroneous fragments of the chromosomal sequences is a nontrivial problem. Minimum error correction approach states to minimize the number of errors to be corrected so that the pair of haplotypes can be constructed through consensus of the fragments. We give a heuristic algorithm that searches through alternative solutions using a gain measure and stops whenever no better solution can be achieved. Time complexity of each iteration is O(m 3 k) for an m × k SNP matrix where m and k are the number of fragments (number of rows) and number of SNP sites (number of columns) respectively in a SNP matrix. Alternative gain measure is also given to reduce running time. Experimental results show that our algorithm outperforms the best known previous algorithm.

International Journal of Parallel Programming, 2015

Random networks are widely used for modeling and analyzing complex processes. Many mathematical m... more Random networks are widely used for modeling and analyzing complex processes. Many mathematical models have been proposed to capture diverse real-world networks. One of the most important aspects of these models is degree distribution. Chung-Lu (CL) model is a random network model, which can produce networks with any given arbitrary degree distribution. The complex systems we deal with nowadays are growing larger and more diverse than ever. Generating random networks with any given degree distribution consisting of billions of nodes and edges or more has become a necessity, which requires efficient and parallel algorithms. We present an MPI-based distributed memory parallel algorithm for generating massive random networks using CL model, which takes O(m+n P + P) time with high probability and O(n) space per processor, where n, m, and P are the number of nodes, edges and processors, respectively. The time efficiency is achieved by using a novel load-balancing algorithm. Our algorithms scale very well to a large number of processors and can generate massive powerlaw networks with one billion nodes and 250 billion edges in one minute using 1024 processors.

NanoBioscience, …, Jan 1, 2012

Clinical symptoms resulting from microbial infection of the gastrointestinal (GI) tract are often... more Clinical symptoms resulting from microbial infection of the gastrointestinal (GI) tract are often exacerbated by inflammation-induced immunopathogenesis. Identifying novel avenues for treating and preventing such pathologies is necessary and complicated by the complexity of interacting immune pathways in the gut, where inflammatory immune cells are regulated by anti-inflammatory cells. The ENteric Immunity Simulator (ENISI) is a simulator of the GI mucosa created for testing and generating hypothesis of host immune mechanisms in response to the presence of resident commensal bacteria and invading pathogens and the effect on host clinical symptoms. ENISI is an implementation of an agent-based model of individual mucosal immune cells each endowed with a program for movement and differentiation according to their cell-type, i.e. epithelial cells, dendritic cells, macrophages, conventional T cells, and natural T-regulatory cells. The internal programs specify movement among the gut lumen, lamina propria, and blood in response to an inflammation-inducing pathogen and tolerance-inducing commensal bacteria. The model focuses on the antagonistic relationship between inflammatory and regulatory (anti-inflammatory) factors whose constant presence characterize mucosal tissue sites.

Parallel & Distributed …, Jan 1, 2012

Here we present the ENteric Immunity Simulator (ENISI), a modeling system for the inflammatory an... more Here we present the ENteric Immunity Simulator (ENISI), a modeling system for the inflammatory and regulatory immune pathways triggered by microbe-immune cell interactions in the gut. With ENISI, immunologists and infectious disease experts can test and generate hypotheses for enteric disease pathology and propose interventions through experimental infection of an in silico gut. ENISI is an agent based simulator, in which individual cells move through the simulated tissues, and engage in context-dependent interactions with the other cells with which they are in contact. The scale of ENISI is unprecedented in this domain, with the ability to simulate 10 7 cells for 250 simulated days on 576 cores in one and a half hours, with the potential to scale to even larger hardware and problem sizes.

differences, Jan 1, 2010

Haplotype is a pattern of SNPs (Single Nucleotide Polymorphism) on a single chromosome. Construct... more Haplotype is a pattern of SNPs (Single Nucleotide Polymorphism) on a single chromosome. Constructing a pair of haplotypes from aligned and overlapping but intermixed and erroneous fragments of the chromosomal sequences is a nontrivial problem. Minimum error correction (MEC) model, which is the mostly used model, minimizes the number of errors to be corrected so that the pair of haplotypes can be constructed through consensus of the fragments. However, this model is effective only when the error rate of SNP fragments is low. To overcome this problem, Zhang et al. proposed a new model called Minimum Conflict Individual Haplotyping (MCIH) as an extension to MEC [1]. This new model uses both SNP fragment information and related genotype information for haplotype reconstruction. MCIH has already been proven to be a potential alternative in individual haplotyping. In this paper, we give a heuristic algorithm for MCIH that searches through alternative solutions using a gain measure and stops whenever no better solution can be achieved. Experimental results on real data show that our algorithm performs better than the best known algorithm for MEC and the algorithm for MCIH proposed by Zhang et al. .

Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Construct... more Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Constructing a pair of haplotypes from aligned and overlapping but intermixedand erroneous fragments of the chromosomal sequences is a nontrivial problem. Minimum error correction approach states to minimize the number of errors to be corrected so that the pair of haplotypes can be constructed through consensus of the fragments. We give a heuristic algorithm that searches through alternative solutions using a gain measure and stops whenever no better solution can be achieved. Time complexity of each iteration is O(m3k) for an m ? k SNP matrix where m and k are the number of fragments (number of rows) and number of SNP sites(number of columns) respectively in a SNP matrix. Alter native gain measure is also given to reduce running time. Experimental results show that our algorithm out performsthe best known previous algorithm.

Data Science and Engineering, 2019

A novel parallel algorithm is presented for generating random scale-free networks using the prefe... more A novel parallel algorithm is presented for generating random scale-free networks using the preferential attachment model. The algorithm, named cuPPA, is custom-designed for "single instruction multiple data (SIMD)" style of parallel processing supported by modern processors such as graphical processing units (GPUs). To the best of our knowledge, our algorithm is the first to exploit GPUs, and also the fastest implementation available today, to generate scale-free networks using the preferential attachment model. A detailed performance study is presented to understand the scalability and runtime characteristics of the cuPPA algorithm. Also another version of the algorithm called cuPPA-Hash tailored for multiple GPUs is presented. On a single GPU, the original cuPPA algorithm delivers the best performance, but is challenging to port to multi-GPU implementation. For multi-GPU implementation, cuPPA-Hash has been used as the parallel algorithm to achieve a perfect linear speedup up to 4 GPUs. In one of the best cases, when executed on an NVidia GeForce 1080 GPU, the original cuPPA generates a scale-free network of two billion edges in less than 3 s. On multi-GPU platforms, cuPPA-Hash generates a scale-free network of 16 billion edges in less than 7 s using a machine consisting of 4 NVidia Tesla P100 GPUs.

Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis, 2013

Recently, there has been substantial interest in the study of various random networks as mathemat... more Recently, there has been substantial interest in the study of various random networks as mathematical models of complex systems. As these complex systems grow larger, the ability to generate progressively large random networks becomes all the more important. This motivates the need for efficient parallel algorithms for generating such networks. Naive parallelization of the sequential algorithms for generating random networks may not work due to the dependencies among the edges and the possibility of creating duplicate (parallel) edges. In this paper, we present MPI-based distributed memory parallel algorithms for generating random scale-free networks using the preferential-attachment model. Our algorithms scale very well to a large number of processors and provide almost linear speedups. The algorithms can generate scale-free networks with 50 billion edges in 123 seconds using 768 processors.

PLoS ONE, 2013

T helper (Th) cells play a major role in the immune response and pathology at the gastric mucosa ... more T helper (Th) cells play a major role in the immune response and pathology at the gastric mucosa during Helicobacter pylori infection. There is a limited mechanistic understanding regarding the contributions of CD4+ T cell subsets to gastritis development during H. pylori colonization. We used two computational approaches: ordinary differential equation (ODE)based and agent-based modeling (ABM) to study the mechanisms underlying cellular immune responses to H. pylori and how CD4+ T cell subsets influenced initiation, progression and outcome of disease. To calibrate the model, in vivo experimentation was performed by infecting C57BL/6 mice intragastrically with H. pylori and assaying immune cell subsets in the stomach and gastric lymph nodes (GLN) on days 0, 7, 14, 30 and 60 post-infection. Our computational model reproduced the dynamics of effector and regulatory pathways in the gastric lamina propria (LP) in silico. Simulation results show the induction of a Th17 response and a dominant Th1 response, together with a regulatory response characterized by high levels of mucosal Treg) cells. We also investigated the potential role of peroxisome proliferator-activated receptor c (PPARc) activation on the modulation of host responses to H. pylori by using loss-of-function approaches. Specifically, in silico results showed a predominance of Th1 and Th17 cells in the stomach of the cell-specific PPARc knockout system when compared to the wild-type simulation. Spatio-temporal, object-oriented ABM approaches suggested similar dynamics in induction of host responses showing analogous T cell distributions to ODE modeling and facilitated tracking lesion formation. In addition, sensitivity analysis predicted a crucial contribution of Th1 and Th17 effector responses as mediators of histopathological changes in the gastric mucosa during chronic stages of infection, which were experimentally validated in mice. These integrated immunoinformatics approaches characterized the induction of mucosal effector and regulatory pathways controlled by PPARc during H. pylori infection affecting disease outcomes.

2012 IEEE 8th International Conference on E-Science, 2012

Networks are an effective abstraction for representing real systems. Consequently, network scienc... more Networks are an effective abstraction for representing real systems. Consequently, network science is increasingly used in academia and industry to solve problems in many fields. Computations that determine structure properties and dynamical behaviors of networks are useful because they give insights into the characteristics of real systems. We introduce a newly built and deployed cyberinfrastructure for network science (CINET) that performs such computations, with the following features: (i) it offers realistic networks from the literature and various random and deterministic network generators; (ii) it provides many algorithmic modules and measures to study and characterize networks; (iii) it is designed for efficient execution of complex algorithms on distributed high performance computers so that they scale to large networks; and (iv) it is hosted with web interfaces so that those without direct access to high performance computing resources and those who are not computing experts can still reap the system benefits. It is a combination of application design and cyberinfrastructure that makes these features possible. To our knowledge, these capabilities collectively make CINET novel. We describe the system and illustrative use cases, with a focus on the CINET user.

2008 International Conference on BioMedical Engineering and Informatics, 2008

Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Construct... more Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Constructing a pair of haplotypes from aligned and overlapping but intermixed and erroneous fragments of the chromosomal sequences is a nontrivial problem. Minimum error correction approach states to minimize the number of errors to be corrected so that the pair of haplotypes can be constructed through consensus of the fragments. We give a heuristic algorithm that searches through alternative solutions using a gain measure and stops whenever no better solution can be achieved. Time complexity of each iteration is O(m 3 k) for an m × k SNP matrix where m and k are the number of fragments (number of rows) and number of SNP sites (number of columns) respectively in a SNP matrix. Alternative gain measure is also given to reduce running time. Experimental results show that our algorithm outperforms the best known previous algorithm.

International Journal of Parallel Programming, 2015

Random networks are widely used for modeling and analyzing complex processes. Many mathematical m... more Random networks are widely used for modeling and analyzing complex processes. Many mathematical models have been proposed to capture diverse real-world networks. One of the most important aspects of these models is degree distribution. Chung-Lu (CL) model is a random network model, which can produce networks with any given arbitrary degree distribution. The complex systems we deal with nowadays are growing larger and more diverse than ever. Generating random networks with any given degree distribution consisting of billions of nodes and edges or more has become a necessity, which requires efficient and parallel algorithms. We present an MPI-based distributed memory parallel algorithm for generating massive random networks using CL model, which takes O(m+n P + P) time with high probability and O(n) space per processor, where n, m, and P are the number of nodes, edges and processors, respectively. The time efficiency is achieved by using a novel load-balancing algorithm. Our algorithms scale very well to a large number of processors and can generate massive powerlaw networks with one billion nodes and 250 billion edges in one minute using 1024 processors.

NanoBioscience, …, Jan 1, 2012

Clinical symptoms resulting from microbial infection of the gastrointestinal (GI) tract are often... more Clinical symptoms resulting from microbial infection of the gastrointestinal (GI) tract are often exacerbated by inflammation-induced immunopathogenesis. Identifying novel avenues for treating and preventing such pathologies is necessary and complicated by the complexity of interacting immune pathways in the gut, where inflammatory immune cells are regulated by anti-inflammatory cells. The ENteric Immunity Simulator (ENISI) is a simulator of the GI mucosa created for testing and generating hypothesis of host immune mechanisms in response to the presence of resident commensal bacteria and invading pathogens and the effect on host clinical symptoms. ENISI is an implementation of an agent-based model of individual mucosal immune cells each endowed with a program for movement and differentiation according to their cell-type, i.e. epithelial cells, dendritic cells, macrophages, conventional T cells, and natural T-regulatory cells. The internal programs specify movement among the gut lumen, lamina propria, and blood in response to an inflammation-inducing pathogen and tolerance-inducing commensal bacteria. The model focuses on the antagonistic relationship between inflammatory and regulatory (anti-inflammatory) factors whose constant presence characterize mucosal tissue sites.

Parallel & Distributed …, Jan 1, 2012

Here we present the ENteric Immunity Simulator (ENISI), a modeling system for the inflammatory an... more Here we present the ENteric Immunity Simulator (ENISI), a modeling system for the inflammatory and regulatory immune pathways triggered by microbe-immune cell interactions in the gut. With ENISI, immunologists and infectious disease experts can test and generate hypotheses for enteric disease pathology and propose interventions through experimental infection of an in silico gut. ENISI is an agent based simulator, in which individual cells move through the simulated tissues, and engage in context-dependent interactions with the other cells with which they are in contact. The scale of ENISI is unprecedented in this domain, with the ability to simulate 10 7 cells for 250 simulated days on 576 cores in one and a half hours, with the potential to scale to even larger hardware and problem sizes.

differences, Jan 1, 2010

Haplotype is a pattern of SNPs (Single Nucleotide Polymorphism) on a single chromosome. Construct... more Haplotype is a pattern of SNPs (Single Nucleotide Polymorphism) on a single chromosome. Constructing a pair of haplotypes from aligned and overlapping but intermixed and erroneous fragments of the chromosomal sequences is a nontrivial problem. Minimum error correction (MEC) model, which is the mostly used model, minimizes the number of errors to be corrected so that the pair of haplotypes can be constructed through consensus of the fragments. However, this model is effective only when the error rate of SNP fragments is low. To overcome this problem, Zhang et al. proposed a new model called Minimum Conflict Individual Haplotyping (MCIH) as an extension to MEC [1]. This new model uses both SNP fragment information and related genotype information for haplotype reconstruction. MCIH has already been proven to be a potential alternative in individual haplotyping. In this paper, we give a heuristic algorithm for MCIH that searches through alternative solutions using a gain measure and stops whenever no better solution can be achieved. Experimental results on real data show that our algorithm performs better than the best known algorithm for MEC and the algorithm for MCIH proposed by Zhang et al. .

Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Construct... more Haplotype is a pattern of Single Nucleotide Polymorphisms (SNP) on a single chromosome. Constructing a pair of haplotypes from aligned and overlapping but intermixedand erroneous fragments of the chromosomal sequences is a nontrivial problem. Minimum error correction approach states to minimize the number of errors to be corrected so that the pair of haplotypes can be constructed through consensus of the fragments. We give a heuristic algorithm that searches through alternative solutions using a gain measure and stops whenever no better solution can be achieved. Time complexity of each iteration is O(m3k) for an m ? k SNP matrix where m and k are the number of fragments (number of rows) and number of SNP sites(number of columns) respectively in a SNP matrix. Alter native gain measure is also given to reduce running time. Experimental results show that our algorithm out performsthe best known previous algorithm.