Pacific Symposium on Biocomputing

Machine learning predictive analytics (MLPA) are utilized increasingly in health care, but can pose harms to patients, clinicians, health systems, and the public. The dynamic nature of this technology creates unique challenges to evaluating safety and efficacy and minimizing harms. In response, regulators have proposed an approach that would shift more responsibility to MLPA developers for mitigating potential harms. To be effective, this approach requires MLPA developers to recognize, accept, and act on responsibility for mitigating harms...

Federated learning is becoming increasingly more popular as the concern of privacy breaches rises across disciplines including the biological and biomedical fields. The main idea is to train models locally on each server using data that are only available to that server and aggregate the model (not data) information at the global level. While federated learning has made significant advancements for machine learning methods such as deep neural networks, to the best of our knowledge, its development in sparse Bayesian models is still lacking...

AI has shown radiologist-level performance at diagnosis and detection of breast cancer from breast imaging such as ultrasound and mammography. Integration of AI-enhanced breast imaging into a radiologist's workflow through the use of computer-aided diagnosis systems, may affect the relationship they maintain with their patient. This raises ethical questions about the maintenance of the radiologist-patient relationship and the achievement of the ethical ideal of shared decision-making (SDM) in breast imaging...

Innovations in human-centered biomedical informatics are often developed with the eventual goal of real-world translation. While biomedical research questions are usually answered in terms of how a method performs in a particular context, we argue that it is equally important to consider and formally evaluate the ethical implications of informatics solutions. Several new research paradigms have arisen as a result of the consideration of ethical issues, including but not limited for privacy-preserving computation and fair machine learning...

Late-onset Alzheimer's disease (LOAD) is a polygenic disorder with a long prodromal phase, making early diagnosis challenging. Twin studies estimate LOAD as 60-80% heritable, and while common genetic variants can account for 30% of this heritability, nearly 70% remains "missing". Polygenic risk scores (PRS) leverage combined effects of many loci to predict LOAD risk, but often lack sensitivity to preclinical disease changes, limiting clinical utility. Our group has built and published on a resilience phenotype to model better-than-expected cognition give amyloid pathology burden and hypothesized it may assist in preclinical polygenic risk prediction...

Polygenic risk scores (PRS) have led to enthusiasm for precision medicine. However, it is well documented that PRS do not generalize across groups differing in ancestry or sample characteristics e.g., age. Quantifying performance of PRS across different groups of study participants, using genome-wide association study (GWAS) summary statistics from multiple ancestry groups and sample sizes, and using different linkage disequilibrium (LD) reference panels may clarify which factors are limiting PRS transferability...

Abdominal aortic aneurysms (AAA) are common enlargements of the abdominal aorta which can grow larger until rupture, often leading to death. Detection of AAA is often by ultrasonography and screening recommendations are mostly directed at men over 65 with a smoking history. Recent large-scale genome-wide association studies have identified genetic loci associated with AAA risk. We combined known risk factors, polygenic risk scores (PRS) and precedent clinical diagnoses from electronic health records (EHR) to develop predictive models for AAA, and compared performance against screening recommendations...

A major goal of precision medicine is to stratify patients based on their genetic risk for a disease to inform future screening and intervention strategies. For conditions like primary open-angle glaucoma (POAG), the genetic risk architecture is complicated with multiple variants contributing small effects on risk. Following the tepid success of genome-wide association studies for high-effect disease risk variant discovery, genetic risk scores (GRS), which collate effects from multiple genetic variants into a single measure, have shown promise for disease risk stratification...

This PSB 2023 session discusses challenges in clinical implication and application of risk prediction models, which includes but is not limited to: implementation of risk models, responsible use of polygenic risk scores (PGS), and other risk prediction strategies. We focus on the development and use of new, scalable methods for harmonizing and refining risk prediction models by incorporating genetic and non-genetic risk factors, applying new phenotyping strategies, and integrating clinical factors and biomarkers...

Deep learning methods for image segmentation and contouring are gaining prominence as an automated approach for delineating anatomical structures in medical images during radiation treatment planning. These contours are used to guide radiotherapy treatment planning, so it is important that contouring errors are flagged before they are used for planning. This creates a need for effective quality assurance methods to enable the clinical use of automated contours in radiotherapy. We propose a novel method for contour quality assurance that requires only shape features, making it independent of the platform used to obtain the images...

As the diversity of genomic variation data increases with our growing understanding of the role of variation in health and disease, it is critical to develop standards for precise inter-system exchange of these data for research and clinical applications. The Global Alliance for Genomics and Health (GA4GH) Variation Representation Specification (VRS) meets this need through a technical terminology and information model for disambiguating and concisely representing variation concepts. Here we discuss the recent Genotype model in VRS, which may be used to represent the allelic composition of a genetic locus...

Preeclampsia is a leading cause of maternal and fetal morbidity and mortality. Currently, the only definitive treatment of preeclampsia is delivery of the placenta, which is central to the pathogenesis of the disease. Transcriptional profiling of human placenta from pregnancies complicated by preeclampsia has been extensively performed to identify differentially expressed genes (DEGs). The decisions to investigate DEGs experimentally are biased by many factors, causing many DEGs to remain uninvestigated. A set of DEGs which are associated with a disease experimentally, but which have no known association to the disease in the literature are known as the ignorome...

We consider the problem of modeling gestational diabetes in a clinical study and develop a domain expert-guided probabilistic model that is both interpretable and explainable. Specifically, we construct a probabilistic model based on causal independence (Noisy-Or) from a carefully chosen set of features. We validate the efficacy of the model on the clinical study and demonstrate the importance of the features and the causal independence model.

Accurate prediction of TCR binding affinity to a target antigen is important for development of immunotherapy strategies. Recent computational methods were built on various deep neural networks and used the evolutionary-based distance matrix BLOSUM to embed amino acids of TCR and epitope sequences to numeric values. A pre-trained language model of amino acids is an alternative embedding method where each amino acid in a peptide is embedded as a continuous numeric vector. Little attention has yet been given to summarize the amino-acid-wise embedding vectors to sequence-wise representations...

The average life expectancy is increasing globally due to advancements in medical technology, preventive health care, and a growing emphasis on gerontological health. Therefore, developing technologies that detect and track aging-associated disease in cognitive function among older adult populations is imperative. In particular, research related to automatic detection and evaluation of Alzheimer's disease (AD) is critical given the disease's prevalence and the cost of current methods. As AD impacts the acoustics of speech and vocabulary, natural language processing and machine learning provide promising techniques for reliably detecting AD...

The accurate interpretation of genetic variants is essential for clinical actionability. However, a majority of variants remain of uncertain significance. Multiplexed assays of variant effects (MAVEs), can help provide functional evidence for variants of uncertain significance (VUS) at the scale of entire genes. Although the systematic prioritization of genes for such assays has been of great interest from the clinical perspective, existing strategies have rarely emphasized this motivation. Here, we propose three objectives for quantifying the importance of genes each satisfying a specific clinical goal: (1) Movability scores to prioritize genes with the most VUS moving to non-VUS categories, (2) Correction scores to prioritize genes with the most pathogenic and/or benign variants that could be reclassified, and (3) Uncertainty scores to prioritize genes with VUS for which variant pathogenicity predictors used in clinical classification exhibit the greatest uncertainty...

Data biases are a known impediment to the development of trustworthy machine learning models and their application to many biomedical problems. When biased data is suspected, the assumption that the labeled data is representative of the population must be relaxed and methods that exploit a typically representative unlabeled data must be developed. To mitigate the adverse effects of unrepresentative data, we consider a binary semi-supervised setting and focus on identifying whether the labeled data is biased and to what extent...

Several biomedical applications contain multiple treatments from which we want to estimate the causal effect on a given outcome. Most existing Causal Inference methods, however, focus on single treatments. In this work, we propose a neural network that adopts a multi-task learning approach to estimate the effect of multiple treatments. We validated M3E2 in three synthetic benchmark datasets that mimic biomedical datasets. Our analysis showed that our method makes more accurate estimations than existing baselines...

A critical challenge in analyzing multi-omics data from clinical cohorts is the re-use of these valuable datasets to answer biological questions beyond the scope of the original study. Transfer Learning and Knowledge Transfer approaches are machine learning methods that leverage knowledge gained in one domain to solve a problem in another. Here, we address the challenge of developing Knowledge Transfer approaches to map trans-omic information from a multi-omic clinical cohort to another cohort in which a novel phenotype is measured...

The discovery of cancer drivers and drug targets are often limited to the biological systems - from cancer model systems to patients. While multiomic patient databases have sparse drug response data, cancer model systems databases, despite covering a broad range of pharmacogenomic platforms, provide lower lineage-specific sample sizes, resulting in reduced statistical power to detect both functional driver genes and their associations with drug sensitivity profiles. Hence, integrating evidence across model systems, taking into account the pros and cons of each system, in addition to multiomic integration, can more efficiently deconvolve cellular mechanisms of cancer as well as learn therapeutic associations...