In the last 15 years, the world has seen a massive digitization of healthcare information driven by the exponential increase and availability of scientific literature, molecular data and real world evidence. Advances in next-generation sequencing have significantly reduced the cost of sequencing and this has enabled large-scale initiatives like the UK Biobank where 500K patients have been sequenced and their de-identified clinico-genomic data is now available to scientists worldwide. Another important driver of this digitization has been the rapid adoption of electronic medical records, which has created an incredible amount of data about human health in the real world. This digitization is affording us an unprecedented opportunity to solve complex medical problems and improve the lives of people on a global scale. However, this explosion of data regarding the biology of disease has far outpaced the human ability to consume and make sense of it. This is a major challenge for scientists working to translate biomedical data into new treatments and diagnostics for patients that need them.
Most of biomedical information is in unstructured form, such as clinical notes, scientific literature, pathology images, ECG waveforms, etc. This information was originally meant for an expert (physician/physician scientist) to interpret. Such unstructured data are rich in biological context and therapeutic outcomes. However, they are not directly computable from a research perspective. The nference technology makes unstructured knowledge computable and enables seamless triangulation with various structured databases that are often siloed (such as vitals, lab tests, ICD codes, genomic sequences). By establishing concordance or discordance of insights computed from differentiated knowledge bases at scale, our platform in turn enables scientists and physicians to answer deeper questions on disease biology, longitudinal progression of treatment outcomes, and therapeutic/diagnostic options to best serve patients.
The specific human cell types that are susceptible to SARS-CoV-2 infection were largely unknown at the onset of the pandemic. Our study was the first to map the viral receptor (ACE2) at unprecedented scale and resolution. This study revealed the gastrointestinal (GI) system and the cardiovascular-renal system as salient foci for COVID-19 pathogenesis beyond the respiratory system.
Molecular mimicry is an evolutionary strategy adopted by viruses to exploit the host machinery. Our study identified that SARS-CoV-2 has evolved a unique FURIN-cleavage site within the viral spike protein, absent in any previous coronavirus sequenced till date. SARS-CoV-2 mimics a FURIN-cleavable peptide on human ENaC-α, which may be associated with the wet lungs syndrome reported in some COVID-19 patients.
In this study, we pioneered the use of deep neural networks for augmented curation of clinical notes from tens of thousands of patients subjected to SARS-CoV-2 PCR testing. Our research revealed that anosmia (altered or loss of sense of smell) and dysgeusia (altered or loss of sense of taste) are the earliest specific indicators of impending COVID-19 diagnosis, up to 4-7 days ahead of the typical PCR testing date.
Understanding the likely course of COVID-19 disease progression requires holistic analysis of all laboratory testing results across thousands of patients. Our study reveals that COVID-19 associated coagulopathy (CAC) is characterised by higher plasma fibrinogen levels and lower platelet counts at the time of clinical presentation, with dramatic changes in these test results during subsequent phases of the disease.
Intensive Care Unit (ICU) admissions and mortality in severe COVID-19 patients are associated with cytokine storms and acute respiratory distress syndrome (ARDS). Our study reveals that plasma IL-6 levels following corticosteroid therapy is a strong indicator of the ICU length of stay (LoS) in critically ill COVID-19 patients.
Our preliminary analysis of over 100,000 patient’s complete immunization records suggests the hand of trained immunity in mitigating the risk of SARS-CoV-2 infection for multiple FDA-approved vaccines. These vaccines are recommended for the majority of the workforce and the pediatric and geriatric populations, thus presenting promising options for amplifying immunity to SARS-CoV-2 infection across age groups.
Our study of longitudinal SARS-CoV-2 PCR testing and the IgG seropositivity assay revealed that a majority of the long-term viral RNA shedders are not hospitalized and are asymptomatic. Our finding that IgG-seropositive individuals can still shed viral RNA emphasizes the need for monitoring viral loads and neutralizing antibody titers in long-term SARS-CoV-2 RNA shedders.
In collaboration with the Mayo Clinic, we have developed and deployed an epidemiological tracking platform that integrates the SARS-CoV-2 PCR testing data at the county level across the United States. This platform enables measuring and predicting spatio-temporal “hot spots” of SARS-CoV-2 infectivity, and has been successfully used to proactively allocate PPE resources.
The Mayo Clinic Platform is a strategic initiative to improve health care through insights and knowledge derived from data. The initial venture under the platform, the Clinical Data Analytics Platform, will apply advanced data analytics on deidentified data from Mayo Clinic and other organizations, as well as the vast information in the scientific literature to advance medicine and improve the health of patients. Mayo Clinic chose nference as its first Clinical Data Analytics Platform partner to accelerate drug discovery and development across the biopharmaceutical ecosystem to create cures for patients. nference will focus on identifying targets and biomarkers for new drugs, optimal matching of patients with therapeutic regimen, and real world data and real world evidence applications, such as label expansion, post-marketing surveillance and drug purposing. nference will be the exclusive partner for analytic services to the biopharmaceutical industry.
nference partners with leading biopharmaceutical companies to solve major challenges in a variety of areas such as drug discovery, clinical research, life cycle management, clinical operations and BD/commercial strategy. As part of these collaborations, nference scientists work closely with experts at those biopharmaceutical companies to answer critical questions by leveraging the nferX software platform and its ability to triangulate insights from both public and proprietary datasets. Learn more about our products and services.
nference is headquartered in Kendall Square (Cambridge, MA), the World’s biotech capital. We believe having a team with deep expertise in technology and biology is one of nference’s main competitive advantages.