Scientific research is increasingly relying on complex workflows that span multiple computing paradigms, including high-performance computing (HPC), high-throughput computing (HTC), and machine learning/artificial intelligence (ML/AI). Traditional monolithic computing infrastructures often struggle to accommodate these diverse and evolving demands. The Reconfigurable Advanced Platform for Transdisciplinary Open Research (RAPTOR) addresses this challenge by providing a dynamically reconfigurable computing environment that integrates with federated resources. RAPTOR's architecture enables dynamic provisioning between an HPC cluster and the Chameleon Cloud platform based on workload requirements, supporting bare-metal customization for specialized applications. This paper focuses on RAPTOR's reconfigurability features and demonstrates their effectiveness through quantitative performance evaluations across four scientific domains: computational proteomics, climate modeling, weather research, and hurricane risk assessment. Our results demonstrate that RAPTOR's reconfigurable design significantly enhances research productivity by providing an appropriate computing environment for diverse computational needs.

Graph Neural Networks (GNNs) have excelled in diverse applications due to their outstanding predictive performance, yet they often overlook fairness considerations, prompting numerous recent efforts to address this societal concern. However, most fair GNNs assume complete demographics by design, which is impractical in most real-world socially sensitive applications due to privacy, legal, or regulatory restrictions. For example, the Consumer Financial Protection Bureau (CFPB) mandates that creditors ensure fairness without requesting or collecting information about an applicant’s race, religion, nationality, sex, or other demographics. To this end, this paper proposes fairGNN-WOD, a first-of-its-kind framework that considers mitigating unfairness in graph learning without using demographic information. In addition, this paper provides a theoretical perspective on analyzing bias in node representations and establishes the relationship between utility and fairness objectives. Experiments on three real-world graph datasets illustrate that fairGNN-WOD outperforms state-of-the-art baselines in achieving fairness but also maintains comparable prediction performance.

Predicting seizures ahead of time will have a significant positive clinical impact for people with epilepsy. Advances in machine learning/artificial intelligence (ML/AI) has provided us the tools needed to perform such predictive tasks. To date, advanced deep learning (DL) architectures such as the convolutional neural network (CNN) and long short-term memory (LSTM) have been used with mixed results. However, highly connected activity exhibited by epileptic seizures necessitates the design of more complex ML techniques which can better capture the complex interconnected neurological processes. Other challenges include the variability of EEG sensor data quality, different epilepsy and seizure profiles, lack of annotated datasets and absence of ML-ready benchmarks. In addition, successful models will need to perform inference in almost real-time using limited hardware compute-capacity. To address these challenges, we propose a lightweight architecture, called ESPFormer, whose novelty lies in the simple and smaller model-size and a lower computational load footprint needed to infer in real-time compared to other works in the literature. To quantify the performance of this lightweight model, we compared its performance with a custom-designed residual neural network (ResNet), a pre-trained vision transformer (ViT) and a pre-trained large-language model (LLM). We tested ESPFormer on MLSPred-Bench which is the largest patient-independent seizure prediction dataset comprising 12 benchmarks. Our results demonstrate that ESPFormer provides the best performance in terms of prediction accuracy for 4/12 benchmarks with an average improvement of 2.65% compared to the LLM, 3.35% compared to the ViT and 17.65% compared to the ResNet – and comparable results for other benchmarks. Our results indicate that lightweight transformer architecture may outperform resource-intensive LLM based models for real-time EEG-based seizure predictions.

Autism Spectrum Disorder (ASD) is a heterogeneous disorder in children, and the current clinical diagnosis is accomplished using behavioral, cognitive, developmental, and language metrics. These clinical metrics can be imperfect measures as they are subject to high test-retest variability, and are influenced by assessment factors such as environment, social structure, or comorbid disorders. Advances in neuroimaging coupled with machine-learning provides an opportunity to develop methods that are more quantifiable, and reliable than existing clinical techniques. In this paper, we design and develop a deep-learning model that operates on functional magnetic resonance imaging (fMRI) data, and can classify between ASD and neurotypical brains. We introduce a novel strategy to transform time-series data extracted from fMRI signals into Gramian Angular Field (GAF) while locking in the temporal and spatial patterns in the data. Our motivation is to design and develop a novel framework that could encode the time-series, acquired from fMRI data, into images that can be used by deep-learning architectures that have been successful in computer vision. In our proposed framework called ASD-GResTM, we used a Convolutional Neural Network (CNN) to extract useful features from GAF images. We then used a Long Short-Term Memory (LSTM) layer to learn the activities between the regions. Finally, the output representations of the last LSTM layer are applied to a single-layer perceptron (SPL) to get the final classification. Our extensive experimentation demonstrates high accuracy across 4 centers, and outperforms state-of-the-art models on two centers with an increase in the accuracy of 17.58% and 6.7%, respectively as compared to the state of the art. Our model achieved the maximum accuracy of 81.78% with high degree of sensitivity and specificity. All training, validation, and testing was accomplished using openly available ABIDE-I benchmarking dataset.

In this paper, we propose an energy-efficient approach for machine-learning based continuous testing and monitoring of long-term patients at the wireless edge. The approach is applicable for any wearable sensors that generate time-series data. Our scheme simultaneously performs sensor-server clustering while ensuring the delay requirements of every user are met. In contrast to previous works on task offloading for generic edge computing/machine learning, our proposed model considers application specific parameters including the sampling rate, measurement duration and number of input channels/leads. We formulate the problem as a mixed integer nonlinear program (MINLP) and propose a heuristic solution. Two applications, cardiac event prediction from a wearable electrocardiograms (ECG), and epileptic seizure prediction from wearable scalp electroencephalography (EEG) are used to demonstrate the superiority of the proposed approach. Results indicate that our proposed algorithm can provide up-to 70% energy savings compared to the case when maximum settings are used.

MOTIVATION: Alzheimer's disease (AD) is a neurodegenerative disease that affects millions of people worldwide. Mild cognitive impairment (MCI) is an intermediary stage between cognitively normal state and AD. Not all people who have MCI convert to AD. The diagnosis of AD is made after significant symptoms of dementia such as short-term memory loss are already present. Since AD is currently an irreversible disease, diagnosis at the onset of the disease brings a huge burden on patients, their caregivers, and the healthcare sector. Thus, there is a crucial need to develop methods for the early prediction AD for patients who have MCI. Recurrent neural networks (RNN) have been successfully used to handle electronic health records (EHR) for predicting conversion from MCI to AD. However, RNN ignores irregular time intervals between successive events which occurs common in electronic health record data. In this study, we propose two deep learning architectures based on RNN, namely Predicting Progression of Alzheimer's Disease (PPAD) and PPAD-Autoencoder. PPAD and PPAD-Autoencoder are designed for early predicting conversion from MCI to AD at the next visit and multiple visits ahead for patients, respectively. To minimize the effect of the irregular time intervals between visits, we propose using age in each visit as an indicator of time change between successive visits. RESULTS: Our experimental results conducted on Alzheimer's Disease Neuroimaging Initiative and National Alzheimer's Coordinating Center datasets showed that our proposed models outperformed all baseline models for most prediction scenarios in terms of F2 and sensitivity. We also observed that the age feature was one of top features and was able to address irregular time interval problem. AVAILABILITY AND IMPLEMENTATION: https://github.com/bozdaglab/PPAD.

Autism spectrum disorder (ASD) affects large number of children and adults in the US, and worldwide. Early and quick diagnosis of ASD can improve the quality of life significantly both for patients and their families. Prior research provides strong evidence that structural and functional magnetic resonance imaging (MRI) data collected from individuals with ASD exhibit distinguishing characteristics that differ in local and global, spatial and temporal neural patterns of the brain - and therefore can be used for diagnostic purposes for various mental disorders. However, the data from MRI are high-dimensional and advanced methods are needed to make sense out of these datasets. In this paper, we present a novel model based on graph convolutional network (GCN) that can utilize resting state fMRI (rs-fMRI) data to classify ASD subjects from health controls (HC). In addition to using the graph from traditional correlation matrices, our proposed GCN model incorporates graphlet topological counting as one of the training features. Our results show that graphlets can preserve the topological information of the graphs obtained from fMRI data. Combined with our GCN, the graphlets retain enough topological information to differentiate between the ASD and HC. Our proposed model gives an average accuracy of 64.27% on the whole ABIDE-I data sets (1035 subjects) and highest site-specific accuracy of 75.9%, which is comparable to other state-of-the-art methods - while potentially open to being more interpretable.

The novel Coronavirus Disease 2019 (COVID-19) is a global pandemic that has infected millions of people causing millions of deaths around the world. Reverse Transcription Polymerase Chain Reaction (RT-PCR) is the standard screening method for COVID-19 detection but it requires specific molecular-biology training. Moreover, the general workflow is difficult e.g. sample collection, processing time, and analysis expertise, etc. Chest radiographic image analysis can be a good alternative screening method that is faster, more efficient, and requires minimal clinical or molecular biology trained laboratory personnel. Early studies have shown that abnormalities on the chest radiographic images are likely to be the consequence of COVID-19 infection. In this study, we propose DeepCOVIDNet, a deep learning based COVID-19 detection model. Our proposed deep-learning model is a multiclass classifier that can distinguish COVID-19, viral pneumonia, bacterial pneumonia, and healthy chest X-ray images. Our proposed model classifies radiographic images into four distinct classes and achieves the accuracy of 89.47% along with a high degree of precision, recall and F1 score. On a different dataset setting (COVID-19, bacterial pneumonia, viral pneumonia) our model achieves the maximum accuracy of 98.25%. We demonstrate generalizability of our proposed method using 5-fold cross validation for COVID-19 vs pneumonia and COVID-19 vs healthy classification that also manifests promising results.

The variability of the results obtained by the statistical analysis of functional human brain networks depend on multiple factors such as: the source of the fMRI data, the brain parcellations, the graph theory measures, and the threshold values applied to the functional connectivity matrices to obtain adjacency matrices of sparse graphs. Therefore, the brain network used for down-stream analysis is heavily dependent on the methods that are applied to the fMRI data to obtain and analyze such networks. In this paper we present the preliminary results of a multi-factorial assessment of the statistical analysis of functional human brain networks. The assessment was performed in the functional human brain networks obtained from the resting state fMRI data of ten imaging sites provided by the Autism Brain Imaging Data Exchange (ABIDE) preprocessed functional magnetic resonance database, with six different functional brain parcellations, six different graph theory measures, and three different threshold values applied to the corresponding connectivity matrices to obtain sparse graphs. The statistical analysis to detect differences between the networks representing autism and control subjects were performed with four different statistical methods, using the p-values to determine the levels of significance of the analysis. Our main results show a strong dependence of functional human brain networks statistical analysis on the brain parcellations, and on the graph theory measures. Our results further show that the results of these analysis are less dependent on the statistical tests methods and on the threshold values of the sparse graphs for all practical purposes. An additional result is that the levels of significance of the statistical tests obtained for data provided by individual sites were much higher than the global levels of significance obtained by averaging the results of all the sites, implying that the best results on the analysis of functional human brain networks are obtained when the source of the fMRI data is the same for all the data. Since reproducibility and reliability of functional brain network statistical analysis is strongly dependent on the graphs obtained from fMRI data; our expectation is that the novel results presented in this paper would further help researchers in this field to develop methods that are reliable and reproducible.

Dissolved organic matter (DOM) is a highly complex mixture of organic substances found in aquatic ecosystems. This mixture results from the degradation of primary producers within the ecosystem, groundwater, and the surrounding terrestrial sources. Understanding the chemical structure of DOM is crucial to assessing its impact on aquatic ecosystems. Although multiple studies have addressed the complexity of DOM, the molecular structure of this set of compounds remains unclear. In this work, we present a novel computational framework "Graph-DOM," to assess the comprehensive fragmentation data obtained from the analysis of DOM using the Data Independent Fragmentation strategy with ESI-FT-ICR MS/MS enabling better understanding of the structural complexity of DOM. Graph-DOM uses graph algorithms to dissect a compiled output file obtained from processing hundreds of ultra-high-resolution fragment spectra. Over half a million ordered fragmentation pathways were computed for 764 isolated precursor ions assuming up to seven vector segments categorized as neutral losses (CH2, CH3, O, CH4, H2O, CO, and CO2). Families of structurally related molecules were identified using pathway overlaps, and output files compatible with network visualization software (e.g., Cytoscape) were also generated. Graph-DOM is able to efficiently process all the pathways to discover families within only a few minutes with adjustable parameters for overlap length of fragmentation pathways as well as configuring low abundance CHOS, CHON, and CHONS compounds. Graph-DOM is available at https://github.com/Usman095/Graph-DOM.

Database algorithms play a crucial part in systems biology studies by identifying proteins from mass spectrometry data. Many of these database search algorithms incur huge computational costs by computing similarity scores for each pair of sparse experimental spectrum and candidate theoretical spectrum vectors. Modern MS instrumentation techniques which are capable of generating high-resolution spectrometry data require comparison against an enormous search space, further emphasizing the need of efficient accelerators. Recent research has shown that the overall cost of scoring, and deducing peptides is dominated by the communication costs between different hierarchies of memory and processing units. However, these communication costs are seldom considered in accelerator-based architectures leading to inefficient DRAM accesses, and poor data-utilization due to irregular memory access patterns. In this paper, we propose a novel communication-avoiding micro-architecture to compute cross-correlation based similarity score by utilizing efficient local cache, and peptide pre-fetching to minimize DRAM accesses, and a custom-designed peptide broadcast bus to allow input reuse. An efficient bus arbitration scheme was designed, and implemented to minimize synchronization cost and exploit parallelism of processing elements. Our simulation results show that the proposed micro-architecture performs on average 24x better than a CPU implementation running on a 3.6 GHz Intel i7-4970 processor with 16GB memory.

In the age of Big Genomics Data, institutions such as the National Human Genome Research Institute (NHGRI) are challenged in their efforts to share volumes of data between researchers, a process that has been plagued by unreliable transfers and slow speeds. These occur due to throughput bottlenecks of traditional transfer technologies. Two factors that affect the efficiency of data transmission are the channel bandwidth and the amount of data. Increasing the bandwidth is one way to transmit data efficiently, but might not always be possible due to resource limitations. Another way to maximize channel utilization is by decreasing the bits needed for transmission of a dataset. Traditionally, transmission of big genomic data between two geographical locations is done using general-purpose protocols, such as hypertext transfer protocol (HTTP) and file transfer protocol (FTP) secure. In this paper, we present a novel deep learning-based data minimization algorithm that 1) minimizes the datasets during transfer over the carrier channels; 2) protects the data from the man-in-the-middle (MITM) and other attacks by changing the binary representation (content-encoding) several times for the same dataset: we assign different codewords to the same character in different parts of the dataset. Our data minimization strategy exploits the alphabet limitation of DNA sequences and modifies the binary representation (codeword) of dataset characters using deep learning-based convolutional neural network (CNN) to ensure a minimum of code word uses to the high frequency characters at different time slots during the transfer time. This algorithm ensures transmission of big genomic DNA datasets with minimal bits and latency and yields an efficient and expedient process. Our tested heuristic model, simulation, and real implementation results indicate that the proposed data minimization algorithm is up to 99 times faster and more secure than the currently used content-encoding scheme used in HTTP of the HTTP content-encoding scheme and 96 times faster than FTP on tested datasets. The developed protocol in C# will be available to the wider genomics community and domain scientists.

Pneumonia is a high mortality disease that kills 50, 000 people in the United States each year. Children under the age of 5 and older population over the age of 65 are susceptible to serious cases of pneumonia. The United States spend billions of dollars fighting pneumonia-related infections every year. Early detection and intervention are crucial in treating pneumonia related infections. Since chest x-ray is one of the simplest and cheapest methods to diagnose pneumonia, we propose a deep learning algorithm based on convolutional neural networks to identify and classify pneumonia cases from these images. For all three models implemented, we obtained varying classification results and accuracy. Based on the results, we obtained better prediction with average accuracy of (68%) and average specificity of (69%) in contrast to the current state-of-the-art accuracy that is (51%) using the Visual Geometry Group (VGG16 also called OxfordNet), which is a convolutional neural network architecture developed by the Visual Geometry Group of Oxford. By implementing more novel lung segmentation techniques, reducing over fitting, and adding more learning layers, the proposed model has the potential to predict at higher accuracy than human specialists and will help subsidies and reduce the cost of diagnosis across the globe.

Database search is the most commonly employed method for identification of peptides from MS/MS spectra data. The search involves comparing experimentally obtained MS/MS spectra against a set of theoretical spectra predicted from a protein sequence database. One of the most commonly employed similarity metrics for spectral comparison is the shared-peak count between a pair of MS/MS spectra. Most modern methods index all generated fragment-ion data from theoretical spectra to speed up the shared peak count computations between a given experimental spectrum and all theoretical spectra. However, the bottleneck for this method is the gigantic memory footprint of fragment-ion index that leads to non-scalable solutions. In this paper, we present a novel data structure, called Compact Fragment-Ion Index Representation (CFIR), that efficiently compresses highly redundant ion-mass information in the data to reduce the index size. Our proposed data structure outperforms all existing fragment-ion indexing data structures by at least 2× in memory consumption while exhibiting the same time complexity for index construction and peptide search. The results also show comparable indexing speed, search speed and speedup scalability for CFIR-index and the state-of-the-art algorithms.

The Sparse Fast Fourier Transform (MIT-SFFT) is an algorithm to compute the discrete Fourier transform of a signal with a sublinear time complexity, i.e. algorithms with runtime complexity proportional to the sparsity level k, where k is the number of non-zero coefficients of the signal in the frequency domain. In this paper, we propose a highly scalable GPU-based parallel algorithm called GPU-SFFT for computing the SFFT of k-sparse signals. Our implementation of GPU-SFFT is based on parallel optimizations that leads to enormous speedups. These include carefully crafting parallel regions in the sequential MIT-SFFT code to exploit parallelism, and minimizing data movement between the CPU and the GPU. This allows us to exploit extreme parallelism for the CPU-GPU architectures and to maximize the number of concurrent threads executing instructions. Our experiments show that our designed CPU-GPU specific optimizations lead to enormous decrease in the run times needed for computing the SFFT. Further we show that GPU-SFFT is 38x times faster than the MIT-SFFT and 5x faster than cuFFT, the NVIDIA CUDA Fast Fourier Transform (FFT) library. The source code for GPU-SFFT is available at https://github.com/pcdslab.

In the field of data science, a huge amount of data, generally represented as graphs, needs to be processed and analyzed. It is of utmost importance that this data be processed swiftly and efficiently to save time and energy. The volume and velocity of data, along with irregular access patterns in graph data structures, pose challenges in terms of analysis and processing. Further, a big chunk of time and energy is spent on analyzing these graphs on large compute clusters and/or data-centers. Filtering and refining of data using graph sampling techniques are one of the most effective ways to speed up the analysis. Efficient accelerators, such as FPGAs, have proven to significantly lower the energy cost of running an algorithm. To this end, we present the design and implementation of a parallel graph sampling technique, for a large number of input graphs streaming into a FPGA. A parallel approach using OpenCL for FPGAs was adopted to come up with a solution that is both time- and energy-efficient. We introduce a novel graph data structure, suitable for streaming graphs on FPGAs, that allows time- and memory-efficient representation of graphs. Our experiments show that our proposed technique is 3x faster and 2x more energy efficient as compared to serial CPU version of the algorithm.

Attention deficit hyperactivity disorder (ADHD) is one of the most common brain disorders among children. This disorder is considered as a big threat for public health and causes attention, focus and organizing difficulties for children and even adults. Since the cause of ADHD is not known yet, data mining algorithms are being used to help discover patterns which discriminate healthy from ADHD subjects. Numerous efforts are underway with the goal of developing classification tools for ADHD diagnosis based on functional and structural magnetic resonance imaging data of the brain. In this paper, we used Eros, which is a technique for computing similarity between two multivariate time series along with k-Nearest-Neighbor classifier, to classify healthy vs ADHD children. We designed a model selection scheme called J-Eros which is able to pick the optimum value of k for k-Nearest-Neighbor from the training data. We applied this technique to the public data provided by ADHD-200 Consortium competition and our results show that J-Eros is capable of discriminating healthy from ADHD children such that we outperformed the best results reported by ADHD-200 competition about 20 percent for two datasets.

It is now possible to compress and decompress large-scale Next-Generation Sequencing files taking advantage of high-performance computing techniques. To this end, we have recently introduced a scalable hybrid parallel algorithm, called phyNGSC, which allows fast compression as well as decompression of big FASTQ datasets using distributed and shared memory programming models via MPI and OpenMP. In this paper we present the design and implementation of a novel parallel data structure which lessens the dependency on decompression and facilitates the handling of DNA sequences in their compressed state using fine-grained decompression in a technique that is identified as in compresso data processing. Using our data structure compression and decompression throughputs of up to 8.71 GB/s and 10.12 GB/s were observed. Our proposed structure and methodology brings us one step closer to compressive genomics and sublinear analysis of big NGS datasets. The code for this implementation is available at https://github.com/pcdslab/PHYNGSD

DNA sequencing has moved into the realm of Big Data due to the rapid development of high-throughput, low cost Next-Generation Sequencing (NGS) technologies. Sequential data compression solutions that once were sufficient to efficiently store and distribute this information are now falling behind. In this paper we introduce phyNGSC, a hybrid MPI-OpenMP strategy to speedup the compression of big NGS data by combining the features of both distributed and shared memory architectures. Our algorithm balances work-load among processes and threads, alleviates memory latency by exploiting locality, and accelerates I/O by reducing excessive read/write operations and inter-node message exchange. To make the algorithm scalable, we introduce a novel timestamp-based file structure that allows us to write the compressed data in a distributed and non-deterministic fashion while retaining the capability of reconstructing the dataset with its original order. Our experimental results show that phyNGSC achieved compression times for big NGS datasets that were 45 to 98 percent faster than NGS-specific sequential compressors with throughputs of up to 3 GB/s. Our theoretical analysis and experimental results suggest strong scalability with some datasets yielding super-linear speedups and constant efficiency. We were able to compress 1 terabyte of data in under 8 minutes compared to more than 5 hours taken by NGS-specific compression algorithms running sequentially. Compared to other parallel solutions, phyNGSC achieved up to 6x speedups while maintaining a higher compression ratio. The code for this implementation is available at https://github.com/pcdslab/PHYNGSC.

MOTIVATION: Modern proteomics studies utilize high-throughput mass spectrometers which can produce data at an astonishing rate. These big mass spectrometry (MS) datasets can easily reach peta-scale level creating storage and analytic problems for large-scale systems biology studies. Each spectrum consists of thousands of peaks which have to be processed to deduce the peptide. However, only a small percentage of peaks in a spectrum are useful for peptide deduction as most of the peaks are either noise or not useful for a given spectrum. This redundant processing of non-useful peaks is a bottleneck for streaming high-throughput processing of big MS data. One way to reduce the amount of computation required in a high-throughput environment is to eliminate non-useful peaks. Existing noise removing algorithms are limited in their data-reduction capability and are compute intensive making them unsuitable for big data and high-throughput environments. In this paper we introduce a novel low-complexity technique based on classification, quantization and sampling of MS peaks. RESULTS: We present a novel data-reductive strategy for analysis of Big MS data. Our algorithm, called MS-REDUCE, is capable of eliminating noisy peaks as well as peaks that do not contribute to peptide deduction before any peptide deduction is attempted. Our experiments have shown up to 100× speed up over existing state of the art noise elimination algorithms while maintaining comparable high quality matches. Using our approach we were able to process a million spectra in just under an hour on a moderate server. AVAILABILITY AND IMPLEMENTATION: The developed tool and strategy has been made available to wider proteomics and parallel computing community and the code can be found at https://github.com/pcdslab/MSREDUCE CONTACT: : [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

High-throughput mass spectrometers can produce massive amounts of redundant data at an astonishing rate with many of them having poor signal-to-noise (S/N) ratio. These low S/N ratio spectra may not get interpreted using conventional spectra-to-database matching techniques. In this paper, we present an efficient algorithm, CAMS-RS (Clustering Algorithm for Mass Spectra using Restricted Space and Sampling) for clustering of raw mass spectrometry data. CAMS-RS utilizes a novel metric (called F-set) that exploits the temporal and spatial patterns to accurately assess similarity between two given spectra. The F-set similarity metric is independent of the retention time and allows clustering of mass spectrometry data from independent LC-MS/MS runs. A novel restricted search space strategy is devised to limit the comparisons of the number of spectra. An intelligent sampling method is executed on individual bins that allow merging of the results to make the final clusters. Our experiments, using experimentally generated data sets, show that the proposed algorithm is able to cluster spectra with high accuracy and is helpful in interpreting low S/N ratio spectra. The CAMS-RS algorithm is highly scalable with increasing number of spectra and our implementation allows clustering of up to a million spectra within minutes.

High-throughput mass spectrometers can produce thousands of peptide spectra from a single complex protein sample in a short amount of time. These data sets contain a substantial amount of redundancy (i.e. the same peptide is selected and identified multiple times in a single experiment) from peptides that may get selected multiple times in the liquid chromatography mass spectrometry (LC-MS/MS) experiment. The data from these mass spectrometers contain a substantial number of spectra that have low signal to noise (S/N) ratio and may not get interpreted due to poor quality. Recently, we presented a graph theoretic algorithm, CAMS (Clustering Algorithm for Mass Spectra) for clustering mass spectrometry data. CAMS utilized a novel metric, called a F-set, that allows accurate identification of the spectra that are similar with much higher accuracy and sensitivity than if single peak comparisons were performed. In this paper we present a multithreaded algorithm, called P-CAMS, for clustering of mass spectral data on multicore machines. The algorithm relies on intelligent matrix completion for graph construction and a load-balancing scheme for substantial speedups. We study the scalability performance of the proposed parallel algorithm on a multicore machine using synthetically generated spectra with parameters carefully chosen to mimic real-world mass spectrometry datasets. Real experimental datasets were also generated for quality assessment of the clustering results from the proposed algorithm. The results show that the proposed algorithms have scalable runtime performances and gives clustering results similar to a serial algorithm. The study also provides insight into the design of high performance algorithms for irregular problems in proteomics on many-core architectures.

High-throughput spectrometers are capable of producing data sets containing thousands of spectra for a single biological sample. These data sets contain a substantial amount of redundancy from peptides that may get selected multiple times in a LC-MS/MS experiment. In this paper, we present an efficient algorithm, CAMS (Clustering Algorithm for Mass Spectra) for clustering mass spectrometry data which increases both the sensitivity and confidence of spectral assignment. CAMS utilizes a novel metric, called F-set, that allows accurate identification of the spectra that are similar. A graph theoretic framework is defined that allows the use of F-set metric efficiently for accurate cluster identifications. The accuracy of the algorithm is tested on real HCD and CID data sets with varying amounts of peptides. Our experiments show that the proposed algorithm is able to cluster spectra with very high accuracy in a reasonable amount of time for large spectral data sets. Thus, the algorithm is able to decrease the computational time by compressing the data sets while increasing the throughput of the data by interpreting low S/N spectra.

Multiple sequence alignment (MSA) is one of the most computationally intensive tasks in Computational Biology. Existing best known solutions for multiple sequence alignment take several hours (in some cases days) of computation time to align, for example, 2000 homologous sequences of average length 300. Inspired by the Sample Sort approach in parallel processing, in this paper we propose a highly scalable multiprocessor solution for the MSA problem in phylogenetically diverse sequences. Our method employs an intelligent scheme to partition the set of sequences into smaller subsets using k- mer count based similarity index, referred to as k-mer rank. Each subset is then independently aligned in parallel using any sequential approach. Further fine tuning of the local alignments is achieved using constraints derived from a global ancestor of the entire set. The proposed sample-align-D algorithm has been implemented on a cluster of workstations using MPI message passing library. The accuracy of the proposed solution has been tested on standard benchmarks such as PREFAB. The accuracy of the alignment produced by our methods is comparable to that of well known sequential MSA techniques. We were able to align 2000 randomly selected sequences from the Methanosarcina acetivorans genome in less than 10 minutes using sample-align-D on a 16 node cluster, compared to over 23 hours on sequential MUSCLE system running on a single cluster node.