Multimodal representation learning for predicting molecule-disease relations.

MOTIVATION: Predicting molecule-disease indications and side effects is important for drug development and pharmacovigilance. Comprehensively mining molecule-molecule, molecule-disease and disease-disease semantic dependencies can potentially improve prediction performance. METHODS: We introduce a Multi-Modal REpresentation Mapping Approach to Predicting molecular-disease relations (M2REMAP) by incorporating clinical semantics learned from electronic health records (EHR) of 12.6 million patients. Specifically, M2REMAP first learns a multimodal molecule representation that synthesizes chemical property and clinical semantic information by mapping molecule chemicals via a deep neural network onto the clinical semantic embedding space shared by drugs, diseases and other common clinical concepts. To infer molecule-disease relations, M2REMAP combines multimodal molecule representation and disease semantic embedding to jointly infer indications and side effects. RESULTS: We extensively evaluate M2REMAP on molecule indications, side effects and interactions. Results show that incorporating EHR embeddings improves performance significantly, for example, attaining an improvement over the baseline models by 23.6% in PRC-AUC on indications and 23.9% on side effects. Further, M2REMAP overcomes the limitation of existing methods and effectively predicts drugs for novel diseases and emerging pathogens. AVAILABILITY AND IMPLEMENTATION: The code is available at https://github.com/celehs/M2REMAP, and prediction results are provided at https://shiny.parse-health.org/drugs-diseases-dev/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Association between acute disease severity and one-year quality of life among post-hospitalisation COVID-19 patients: Coalition VII prospective cohort study.

PURPOSE: To assess the association between acute disease severity and 1-year quality of life in patients discharged after hospitalisation due to coronavirus disease 2019 (COVID-19). METHODS: We conducted a prospective cohort study nested in 5 randomised clinical trials between March 2020 and March 2022 at 84 sites in Brazil. Adult post-hospitalisation COVID-19 patients were followed for 1 year. The primary outcome was the utility score of EuroQol five-dimension three-level (EQ-5D-3L). Secondary outcomes included all-cause mortality, major cardiovascular events, and new disabilities in instrumental activities of daily living. Adjusted generalised estimating equations were used to assess the association between outcomes and acute disease severity according to the highest level on a modified ordinal scale during hospital stay (2: no oxygen therapy; 3: oxygen by mask or nasal prongs; 4: high-flow nasal cannula oxygen therapy or non-invasive ventilation; 5: mechanical ventilation). RESULTS: 1508 COVID-19 survivors were enrolled. Primary outcome data were available for 1156 participants. At 1 year, compared with severity score 2, severity score 5 was associated with lower EQ-5D-3L utility scores (0.7 vs 0.84; adjusted difference, - 0.1 [95% CI - 0.15 to - 0.06]); and worse results for all-cause mortality (7.9% vs 1.2%; adjusted difference, 7.1% [95% CI 2.5%-11.8%]), major cardiovascular events (5.6% vs 2.3%; adjusted difference, 2.6% [95% CI 0.6%-4.6%]), and new disabilities (40.4% vs 23.5%; adjusted difference, 15.5% [95% CI 8.5%-22.5]). Severity scores 3 and 4 did not differ consistently from score 2. CONCLUSIONS: COVID-19 patients who needed mechanical ventilation during hospitalisation have lower 1-year quality of life than COVID-19 patients who did not need mechanical ventilation during hospitalisation.

Multiview Incomplete Knowledge Graph Integration with application to cross-institutional EHR data harmonization.

OBJECTIVE: The growing availability of electronic health records (EHR) data opens opportunities for integrative analysis of multi-institutional EHR to produce generalizable knowledge. A key barrier to such integrative analyses is the lack of semantic interoperability across different institutions due to coding differences. We propose a Multiview Incomplete Knowledge Graph Integration (MIKGI) algorithm to integrate information from multiple sources with partially overlapping EHR concept codes to enable translations between healthcare systems. METHODS: The MIKGI algorithm combines knowledge graph information from (i) embeddings trained from the co-occurrence patterns of medical codes within each EHR system and (ii) semantic embeddings of the textual strings of all medical codes obtained from the Self-Aligning Pretrained BERT (SAPBERT) algorithm. Due to the heterogeneity in the coding across healthcare systems, each EHR source provides partial coverage of the available codes. MIKGI synthesizes the incomplete knowledge graphs derived from these multi-source embeddings by minimizing a spherical loss function that combines the pairwise directional similarities of embeddings computed from all available sources. MIKGI outputs harmonized semantic embedding vectors for all EHR codes, which improves the quality of the embeddings and enables direct assessment of both similarity and relatedness between any pair of codes from multiple healthcare systems. RESULTS: With EHR co-occurrence data from Veteran Affairs (VA) healthcare and Mass General Brigham (MGB), MIKGI algorithm produces high quality embeddings for a variety of downstream tasks including detecting known similar or related entity pairs and mapping VA local codes to the relevant EHR codes used at MGB. Based on the cosine similarity of the MIKGI trained embeddings, the AUC was 0.918 for detecting similar entity pairs and 0.809 for detecting related pairs. For cross-institutional medical code mapping, the top 1 and top 5 accuracy were 91.0% and 97.5% when mapping medication codes at VA to RxNorm medication codes at MGB; 59.1% and 75.8% when mapping VA local laboratory codes to LOINC hierarchy. When trained with 500 labels, the lab code mapping attained top 1 and 5 accuracy at 77.7% and 87.9%. MIKGI also attained best performance in selecting VA local lab codes for desired laboratory tests and COVID-19 related features for COVID EHR studies. Compared to existing methods, MIKGI attained the most robust performance with accuracy the highest or near the highest across all tasks. CONCLUSIONS: The proposed MIKGI algorithm can effectively integrate incomplete summary data from biomedical text and EHR data to generate harmonized embeddings for EHR codes for knowledge graph modeling and cross-institutional translation of EHR codes.

A Phenome-Wide Association Study of genes associated with COVID-19 severity reveals shared genetics with complex diseases in the Million Veteran Program.

The study aims to determine the shared genetic architecture between COVID-19 severity with existing medical conditions using electronic health record (EHR) data. We conducted a Phenome-Wide Association Study (PheWAS) of genetic variants associated with critical illness (n = 35) or hospitalization (n = 42) due to severe COVID-19 using genome-wide association summary data from the Host Genetics Initiative. PheWAS analysis was performed using genotype-phenotype data from the Veterans Affairs Million Veteran Program (MVP). Phenotypes were defined by International Classification of Diseases (ICD) codes mapped to clinically relevant groups using published PheWAS methods. Among 658,582 Veterans, variants associated with severe COVID-19 were tested for association across 1,559 phenotypes. Variants at the ABO locus (rs495828, rs505922) associated with the largest number of phenotypes (nrs495828 = 53 and nrs505922 = 59); strongest association with venous embolism, odds ratio (ORrs495828 1.33 (p = 1.32 x 10-199), and thrombosis ORrs505922 1.33, p = 2.2 x10-265. Among 67 respiratory conditions tested, 11 had significant associations including MUC5B locus (rs35705950) with increased risk of idiopathic fibrosing alveolitis OR 2.83, p = 4.12 × 10-191; CRHR1 (rs61667602) associated with reduced risk of pulmonary fibrosis, OR 0.84, p = 2.26× 10-12. The TYK2 locus (rs11085727) associated with reduced risk for autoimmune conditions, e.g., psoriasis OR 0.88, p = 6.48 x10-23, lupus OR 0.84, p = 3.97 x 10-06. PheWAS stratified by ancestry demonstrated differences in genotype-phenotype associations. LMNA (rs581342) associated with neutropenia OR 1.29 p = 4.1 x 10-13 among Veterans of African and Hispanic ancestry but not European. Overall, we observed a shared genetic architecture between COVID-19 severity and conditions related to underlying risk factors for severe and poor COVID-19 outcomes. Differing associations between genotype-phenotype across ancestries may inform heterogenous outcomes observed with COVID-19. Divergent associations between risk for severe COVID-19 with autoimmune inflammatory conditions both respiratory and non-respiratory highlights the shared pathways and fine balance of immune host response and autoimmunity and caution required when considering treatment targets.

Pharmacoepidemiology, Machine Learning, and COVID-19: An Intent-to-Treat Analysis of Hydroxychloroquine, With or Without Azithromycin, and COVID-19 Outcomes Among Hospitalized US Veterans.

Hydroxychloroquine (HCQ) was proposed as an early therapy for coronavirus disease 2019 (COVID-19) after in vitro studies indicated possible benefit. Previous in vivo observational studies have presented conflicting results, though recent randomized clinical trials have reported no benefit from HCQ among patients hospitalized with COVID-19. We examined the effects of HCQ alone and in combination with azithromycin in a hospitalized population of US veterans with COVID-19, using a propensity score-adjusted survival analysis with imputation of missing data. According to electronic health record data from the US Department of Veterans Affairs health care system, 64,055 US Veterans were tested for the virus that causes COVID-19 between March 1, 2020 and April 30, 2020. Of the 7,193 veterans who tested positive, 2,809 were hospitalized, and 657 individuals were prescribed HCQ within the first 48-hours of hospitalization for the treatment of COVID-19. There was no apparent benefit associated with HCQ receipt, alone or in combination with azithromycin, and there was an increased risk of intubation when HCQ was used in combination with azithromycin (hazard ratio = 1.55; 95% confidence interval: 1.07, 2.24). In conclusion, we assessed the effectiveness of HCQ with or without azithromycin in treatment of patients hospitalized with COVID-19, using a national sample of the US veteran population. Using rigorous study design and analytic methods to reduce confounding and bias, we found no evidence of a survival benefit from the administration of HCQ.

Phenome-wide association of 1809 phenotypes and COVID-19 disease progression in the Veterans Health Administration Million Veteran Program.

BACKGROUND: The risk factors associated with the stages of Coronavirus Disease-2019 (COVID-19) disease progression are not well known. We aim to identify risk factors specific to each state of COVID-19 progression from SARS-CoV-2 infection through death. METHODS AND RESULTS: We included 648,202 participants from the Veteran Affairs Million Veteran Program (2011-). We identified characteristics and 1,809 ICD code-based phenotypes from the electronic health record. We used logistic regression to examine the association of age, sex, body mass index (BMI), race, and prevalent phenotypes to the stages of COVID-19 disease progression: infection, hospitalization, intensive care unit (ICU) admission, and 30-day mortality (separate models for each). Models were adjusted for age, sex, race, ethnicity, number of visit months and ICD codes, state infection rate and controlled for multiple testing using false discovery rate (≤0.1). As of August 10, 2020, 5,929 individuals were SARS-CoV-2 positive and among those, 1,463 (25%) were hospitalized, 579 (10%) were in ICU, and 398 (7%) died. We observed a lower risk in women vs. men for ICU and mortality (Odds Ratio (95% CI): 0.48 (0.30-0.76) and 0.59 (0.31-1.15), respectively) and a higher risk in Black vs. Other race patients for hospitalization and ICU (OR (95%CI): 1.53 (1.32-1.77) and 1.63 (1.32-2.02), respectively). We observed an increased risk of all COVID-19 disease states with older age and BMI ≥35 vs. 20-24 kg/m2. Renal failure, respiratory failure, morbid obesity, acid-base balance disorder, white blood cell diseases, hydronephrosis and bacterial infections were associated with an increased risk of ICU admissions; sepsis, chronic skin ulcers, acid-base balance disorder and acidosis were associated with mortality. CONCLUSIONS: Older age, higher BMI, males and patients with a history of respiratory, kidney, bacterial or metabolic comorbidities experienced greater COVID-19 severity. Future studies to investigate the underlying mechanisms associated with these phenotype clusters and COVID-19 are warranted.

Quality of life and long-term outcomes after hospitalization for COVID-19: Protocol for a prospective cohort study (Coalition VII). / Qualidade de vida e desfechos em longo prazo após hospitalização por COVID-19: Protocolo para um estudo de coorte prospectivo (Coalizão VII).

INTRODUCTION: The long-term effects caused by COVID-19 are unknown. The present study aims to assess factors associated with health-related quality of life and long-term outcomes among survivors of hospitalization for COVID-19 in Brazil. METHODS: This is a multicenter prospective cohort study nested in five randomized clinical trials designed to assess the effects of specific COVID-19 treatments in over 50 centers in Brazil. Adult survivors of hospitalization due to proven or suspected SARS-CoV-2 infection will be followed-up for a period of 1 year by means of structured telephone interviews. The primary outcome is the 1-year utility score of health-related quality of life assessed by the EuroQol-5D3L. Secondary outcomes include all-cause mortality, major cardiovascular events, rehospitalizations, return to work or study, physical functional status assessed by the Lawton-Brody Instrumental Activities of Daily Living, dyspnea assessed by the modified Medical Research Council dyspnea scale, need for long-term ventilatory support, symptoms of anxiety and depression assessed by the Hospital Anxiety and Depression Scale, symptoms of posttraumatic stress disorder assessed by the Impact of Event Scale-Revised, and self-rated health assessed by the EuroQol-5D3L Visual Analog Scale. Generalized estimated equations will be performed to test the association between five sets of variables (1- demographic characteristics, 2- premorbid state of health, 3- characteristics of acute illness, 4- specific COVID-19 treatments received, and 5- time-updated postdischarge variables) and outcomes. ETHICS AND DISSEMINATION: The study protocol was approved by the Research Ethics Committee of all participant institutions. The results will be disseminated through conferences and peer-reviewed journals.

INTRODUÇÃO: Os efeitos provocados pela COVID-19 em longo prazo são desconhecidos. O presente estudo tem como objetivo avaliar os fatores associados com a qualidade de vida relacionada à saúde e os desfechos em longo prazo em sobreviventes à hospitalização por COVID-19 no Brasil. MÉTODOS: Este será um estudo multicêntrico de coorte prospectivo, aninhado em cinco ensaios clínicos randomizados desenhados para avaliar os efeitos dos tratamentos específicos para COVID-19 em mais de 50 centros no Brasil. Pacientes adultos sobreviventes à hospitalização por infecção por SARS-CoV-2 comprovada ou suspeita serão seguidos por um período de 1 ano, por meio de entrevistas telefônicas estruturadas. O desfecho primário é o escore de utilidade para qualidade de vida relacionada à saúde após 1 ano, avaliado segundo o questionário EuroQol-5D3L. Os desfechos secundários incluirão mortalidade por todas as causas, eventos cardiovasculares graves, reospitalizações, retorno ao trabalho ou estudo, condição funcional física avaliada pelo instrumento Lawton-Brody Instrumental Activities of Daily Living, dispneia avaliada segundo a escala de dispneia modificada do Medical Research Council, necessidade de suporte ventilatório em longo prazo, sintomas de ansiedade e depressão avaliados segundo a Hospital Anxiety and Depression Scale, sintomas de transtorno de estresse pós-traumático avaliados pela ferramenta Impact of Event Scale-Revised e autoavaliação da condição de saúde, conforme a Escala Visual Analógica do EuroQol-5D3L. Serão utilizadas equações de estimativas generalizada para testar a associação entre cinco conjuntos de variáveis (1 - características demográficas, 2 - condição de saúde pré-morbidade, 3 - características da doença aguda, 4 - terapias específicas para COVID-19 recebidas e 5 - variáveis pós-alta atualizadas) e desfechos. ÉTICA E DISSEMINAÇÃO: O protocolo do estudo foi aprovado pelos Comitês de Ética em Pesquisa de todas as instituições participantes. Os resultados serão disseminados por meio de conferências e periódicos revisados por pares.