Deep Learning Algorithms for Secure and Efficient Compression of Genomic Sequence Data

Pathogen genomic data analysis can be extremely bespoke and diverse. This paper presents our plan and progress towards creating a Scalable Pathogen Pipeline Platform (SP 3 ) providing an efficient and unified process of collecting, analysing and comparing genomic data analysis with the benefit of elastic cloud computing. SP 3 enables container-centric bioinformatic workflows run on personal computers, High-performance computing (HPC) clusters and cloud platforms. We have deployed and tested SP 3 on local HPC, Google Cloud Platform (GCP), Microsoft Azure and OpenStack Platforms. SP 3 allows users to fetch genomic sequencing data from European Nucleotide Archive (ENA) and conduct analysis with open-source bioinformatic pipelines. We believe SP 3 will promote common standards around pathogen genomic data quality, data processing and data analysis, helping answer the challenges of tools divergence and leveraging a pool of public genomic data repository and cloud resources.

US-EU Scientific Research Collaborations: Sage Bionetworks’ Experience Navigating the Complex Regulatory Landscape.

The convergence of quantum computing and artificial intelligence (AI) presents a paradigm shift in healthcare, revolutionizing complex biological simulations, genomic data processing, and drug discovery innovations. Traditional computational methods, despite their advancements, often struggle with the sheer scale and complexity of biological data, limiting the speed and accuracy of medical breakthroughs. Quantum computing, with its ability to process vast datasets exponentially faster than classical computers, coupled with AI's predictive capabilities, offers a transformative solution for accelerating biomedical research and clinical applications. This paper explores quantum machine learning’s role in optimizing AI-driven molecular dynamics simulations for drug discovery. By leveraging quantum-enhanced algorithms, researchers can rapidly model molecular interactions, analyze drug-receptor binding affinities, and predict pharmacokinetics with unprecedented precision. Additionally, we examine quantum-assisted deep learning models for deciphering intricate biological mechanisms such as protein folding, epigenetic modifications, and metabolic pathway interactions, enabling more accurate predictions of disease progression and therapeutic targets. Furthermore, the integration of AI-quantum hybrid models in clinical diagnostics and imaging analytics is redefining personalized medicine. Quantum-enhanced deep learning facilitates high-resolution medical imaging, real-time anomaly detection, and optimized radiomic feature extraction, leading to early and more accurate disease diagnosis. In genomics, quantum computing significantly accelerates whole-genome sequencing and mutation analysis, paving the way for tailored treatment strategies based on an individual's genetic profile. Despite its promise, challenges such as quantum hardware limitations, data coherence issues, and ethical considerations must be addressed to ensure the practical implementation of quantum-AI healthcare solutions. This paper provides a comprehensive analysis of the potential, challenges, and future directions of quantum-AI synergy in transforming modern healthcare.

Introduction: The evolution of precision oncology will demand the integration of multi-omics, clinical data, demographics and outcomes, which is a complex and computationally intensive endeavor. Accurate and robust molecular characterization of pediatric leukemias require a comprehensive approach involving longitudinal sample collection across a large cohort. Genomic alterations, both germline and somatic, need to be efficiently captured and annotated to identify clinically meaningful trends. Algorithms developed for conducting genomic analysis typically leverage a central processing unit (CPU) environment for data processing and traditional statistical approach, such as Bayes' theorem, for variant detection. More recently, algorithms have been developed that can leverage a graphics processing unit (GPU) architecture and machine learning (ML) technology. In this study we benchmarked these techniques and applied them sequencing data generated from pediatric AML subjects enrolled on the Children's Oncology Group AAML1031 trial (951 samples). Methods: Whole genome sequencing was conducted using the Illumina platform (2X150) and sequenced between 30x-60x coverage. Genome alignment files were generated and subjected to germline and somatic variant calling (NVIDIA's Selene DGXA100 cluster; Parabricks Pipelines). Germline variants were called using Haplotype caller and DeepVariant (Parabricks v3.0). HG002 genome in a bottle sample was used as a positive control to determine precision, recall and F1 score. Somatic variant calling was completed using somatic sniper, VarScan2, Muse, MuTect2, and Strelka (Parabricks v3.5) and SEQCII data was used to assess somatic variant calling performance. Results: Control sample HG002 was used for benchmarking the algorithms tested. At all depths analyzed, 50X, 30X, 15X, and 5X, GoogleDeep Variant had the highest F1 score, precision, and recall compared to Haplotyper for SNPs and InDels. For somatic SNV detection using SEQCII dataset, Strelka2 had the highest F1 score, precision, and recall (0.9435, 0.9445, 0.9425), followed by MuTect2, MUSE, VarScan2, and SomaticSniper. For somatic InDel detection, MuTect2 and Strelka2 performed similar with MuTect2 higher precision, but Strelka2 had higher recall. These pipelines were utilized to analyze 951 pediatric AML samples, and demonstrated superior processing time and identification of variants, both germline and somatic. Discussion: ML algorithms integrated into a GPU computing environment demonstrated greater precision and accuracy for germline variant detection compared to traditional statistical approaches for variant detection. Ongoing efforts are focused at integration of multiple data sources correlated with outcomes and analyzed via ML tools in order to improve future risk stratification and therapeutic selection for superior outcomes. Citation Format: Erin L. Crowgey, Pankaj Vats, Karl Franke, Gary Burnett, Ankit Sethia, Timothy Harkins, Todd E. Druley. Enhanced processing of genomic sequencing data for pediatric cancers: GPUs and machine learning techniques for variant detection [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 165.

In recent years, the widespread utilization of biological data processing technology has been driven by its cost-effectiveness. Consequently, next-generation sequencing (NGS) has become an integral component of biological research. NGS technologies enable the sequencing of billions of nucleotides in the entire genome, transcriptome, or specific target regions. This sequencing generates vast data matrices. Consequently, there is a growing demand for deep learning (DL) approaches, which employ multilayer artificial neural networks and systems capable of extracting meaningful information from these extensive data structures. In this study, the aim was to obtain optimized parameters and assess the prediction performance of deep learning and machine learning (ML) algorithms for binary classification in real and simulated whole genome data using a cloud-based system. The ART-simulated data and paired-end NGS (whole genome) data of Ch22, which includes ethnicity information, were evaluated using XGBoost, LightGBM, and DL algorithms. When the learning rate was set to 0.01 and 0.001, and the epoch values were updated to 500, 1000, and 2000 in the deep learning model for the ART simulated dataset, the median accuracy values of the ART models were as follows: 0.6320, 0.6800, and 0.7340 for epoch 0.01; and 0.6920, 0.7220, and 0.8020 for epoch 0.001, respectively. In comparison, the median accuracy values of the XGBoost and LightGBM models were 0.6990 and 0.6250 respectively. When the same process is repeated for Chr 22, the results are as follows: the median accuracy values of the DL models were 0.5290, 0.5420 and 0.5820 for epoch 0.01; and 0.5510, 0.5830 and 0.6040 for epoch 0.001, respectively. Additionally, the median accuracy values of the XGBoost and LightGBM models were 0.5760 and 0.5250, respectively. While the best classification estimates were obtained at 2000 epochs and a learning rate (LR) value of 0.001 for both real and simulated data, the XGBoost algorithm showed higher performance when the epoch value was 500 and the LR was 0.01. When dealing with class imbalance, the DL algorithm yielded similar and high Recall and Precision values. Conclusively, this study serves as a timely resource for genomic scientists, providing guidance on why, when, and how to effectively utilize deep learning/machine learning methods for the analysis of human genomic data.

Integrating artificial intelligence (AI), particularly large language models (LLMs), into the healthcare industry is revolutionizing the field of medicine. LLMs possess the capability to analyze the scientific literature and genomic data by comprehending and producing human-like text. This enhances the accuracy, precision, and efficiency of extensive genomic analyses through contextualization. LLMs have made significant advancements in their ability to understand complex genetic terminology and accurately predict medical outcomes. These capabilities allow for a more thorough understanding of genetic influences on health issues and the creation of more effective therapies. This review emphasizes LLMs' significant impact on healthcare, evaluates their triumphs and limitations in genomic data processing, and makes recommendations for addressing these limitations in order to enhance the healthcare system. It explores the latest advancements in LLMs for genomic analysis, focusing on enhancing disease diagnosis and treatment accuracy by taking into account an individual's genetic composition. It also anticipates a future in which AI-driven genomic analysis is commonplace in clinical practice, suggesting potential research areas. To effectively leverage LLMs' potential in personalized medicine, it is vital to actively support innovation across multiple sectors, ensuring that AI developments directly contribute to healthcare solutions tailored to individual patients.

Health informatics (HI) has become a significant research area due to the massive generation of digital health and medical data by biomedical and health research organizations. The health data sources are available in different forms namely electronic health records (EHRs), biomedical imaging, bio-signals, sensor data, genomic data, medical history, social media data, and so on. The structured health data can be utilized for HI and effective predictive modeling of health data assists in the decision-making process. The recently developed artificial intelligence (AI), machine learning (ML), and deep learning (DL) techniques pave a way for effective predictive modeling on health data. Numerous existing works have been presented in the literature depending upon the ML and DL based HI for various applications. With this motivation, this study aims to review the recent state of art ML and DL based predictive models for health sector. This survey primarily identifies the difference between the ML and DL architectures with their significance in health sector. In addition, the existing works are extensively reviewed and compared in terms of different aspects such as objectives, underlying methodology, input source, dataset used, performance validation, metrics, and so on. Finally, the open challenges and future scope of the HI are examined in detail. At the end of the survey, the readers find it useful to identify the present research and possible future scope of the ML and DL based predictive models for HI.

Geneticists prefer to store patients’ aligned, raw genomic data, in addition to their variant calls (compact and summarized form of the raw data), mainly because of the immaturity of bioinformatic algorithms and sequencing platforms. Thus, we propose a privacy-preserving system to protect the privacy of aligned, raw genomic data. The raw genomic data of a patient includes millions of short reads, each comprised of between 100 and 400 nucleotides (genomic letters). We propose storing these short reads at a biobank in encrypted form. The proposed scheme enables a medical unit (e.g., a pharmaceutical company or a hospital) to privately retrieve a subset of the short reads of the patients (which include a definite range of nucleotides depending on the type of the genetic test) without revealing the nature of the genetic test to the biobank. Furthermore, the proposed scheme lets the biobank mask particular parts of the retrieved short reads if (i) some parts of the provided short reads are out of the requested range, or (ii) the patient does not give consent to some parts of the provided short reads (e.g., parts revealing sensitive diseases). We evaluate the proposed scheme to show the amount of unauthorized genomic data leakage it prevents. Finally, we implement the proposed scheme and assess its practicality.KeywordsGenomicsPrivacyBioinformaticsRaw genomic data

As the amount of re-sequencing genome data grows, minimizing the execution time of an analysis is required. For this purpose, recent computing systems have been adopting both high-performance coprocessors and host processors. However, there are few applications that efficiently utilize these heterogeneous computing resources. This problem equally refers to the work of single nucleotide polymorphism (SNP) detection, which is one of the bottlenecks in genome data processing. In this paper, we propose a method for speeding up an SNP detection by enhancing the utilization of heterogeneous computing resources often used in recent high-performance computing systems. Through the measurement of workload in the detection procedure, we divide the SNP detection into several task groups suitable for each computing resource. These task groups are scheduled using a window overlapping method. As a result, we improved upon the speedup achieved by previous open source applications by a magnitude of 10.

Privacy-preserving data publishing is a mechanism for sharing data while ensuring the privacy of individuals is preserved in the published data and utility is maintained for data mining and analysis. There is a huge need for sharing genomic data to advance medical and health research. However, since genomic data is highly sensitive and the ultimate identifier, it is a big challenge to publish genomic data while protecting the privacy of individuals in the data. In this thesis, we address the aforementioned challenge by presenting an approach for privacy-preserving genomic data publishing via differentially-private suffix tree. The proposed algorithm uses a top-down approach and utilizes Laplace mechanism to divide the raw genomic data into disjoint partitions, and then normalize the partitioning structure to ensure consistency and maintain utility. The output of our algorithm is a differentially-private suffix tree, a data structure most suitable for efficient search on genomic data. We experiment on real-life genomic data obtained from the Human Genome Privacy Challenge project, and we show that our approach is efficient, scalable, and achieves high utility with respect to genomic sequence matching count queries.

Australian public perspectives on genomic data storage and sharing: Benefits, concerns and access preferences

The advent of big data in genomics has transformed the landscape of biological research and personalized medicine, offering unprecedented insights into human genetics, disease mechanisms, and treatment development. However, the exponential growth of genomic data presents significant challenges in data storage, processing, analysis, and interpretation. This paper explores the key issues faced in managing and analyzing vast genomic datasets, including data heterogeneity, computational requirements, and privacy concerns. It also discusses emerging solutions and technologies, such as cloud computing, artificial intelligence, and advanced algorithms, that promise to harness the full potential of big data in genomics. By addressing these challenges, the field can unlock new opportunities for breakthroughs in medical research, precision medicine, and public health.

<span>The 21st centuries were deemed to be the era of big data. Data driven research had become a necessity. This hold true not only in the business world, yet also in the field of biomedical world. From a few years of biological data extraction and derivation. With the advancement of Next Generation Sequencing, genomics data had grown to become an ambiguous giant which could not keep up with the pace of its advancement in it analysis counter parts. This results in a large amount of unanalysed genomic data. These genomic data consist not only plain information, researcher had discovered the potential of most gene called the non-coding variant and still failing in identifying their function. With the growth in volume of data, there is also a growth of hardware or technologies. With current technologies, we were able to implement a more complex and sophisticated algorithm in analysis these genomics data. The domain of deep learning had become a major interest of researcher as it was proven to have achieve a significant success in deriving insight from various field. This paper aims to review the current trend of non-coding variant analysis using deep learning approach.</span>

New Technologies, Tools and Approaches for Improving Crop Breeding

Predicting disease status for a complex human disease using genomic data is an important, yet challenging, step in personalized medicine. Among many challenges, the so-called curse of dimensionality problem results in unsatisfied performances of many state-of-art machine learning algorithms. A major recent advance in machine learning is the rapid development of deep learning algorithms that can efficiently extract meaningful features from high-dimensional and complex datasets through a stacked and hierarchical learning process. Deep learning has shown breakthrough performance in several areas including image recognition, natural language processing, and speech recognition. However, the performance of deep learning in predicting disease status using genomic datasets is still not well studied. In this article, we performed a review on the four relevant articles that we found through our thorough literature review. All four articles used auto-encoders to project high-dimensional genomic data to a low dimensional space and then applied the state-of-the-art machine learning algorithms to predict disease status based on the low-dimensional representations. This deep learning approach outperformed existing prediction approaches, such as prediction based on probe-wise screening and prediction based on principal component analysis. The limitations of the current deep learning approach and possible improvements were also discussed.