DILIrank: the largest reference drug list ranked by the risk for developing drug-induced liver injury in humans.

Introduction Rapid advancement in high-throughput technologies, such as high-content assay/high-throughput screening methodologies, enables the assessment of cellular responses to hundreds of drugs in a single experiment. In addition, new approaches utilizing in vitro models (e.g., 3D cell culture), ‘omics’, organ-on-a-chip, and pluripotent stem cells (iPSC) have been introduced to evaluate drug safety and efficacy through anchoring phenotypes observed in humans [1,2]. In this era of data-driven science, analysis anchored in phenotype depends on accurate and consistent annotation for a large number of drugs in a comparative analysis to derive reliable and robust emerging biomarkers. Drug-induced liver injury (DILI) is one of the major safety concerns for drug developers, regulators, and clinicians [3,4]. Current in vivo toxicological studies are not sufficient in assessing the hepatotoxic potential of a drug in humans, as suggested by a large survey conducted by collaborative efforts [5]. Thus, active research is conducted for developing new tools and approaches to better predict DILI risk in humans [6]. A reference drug list with a sufficient number of drugs that are well annotated based on their DILI risk in humans [7] is required for enhanced methodological developments in DILI risk assessment. The DILI annotation addressed here refers to the classification of drugs based on DILI risk for humans treated for various diseases. Establishing a DILI annotation, which reflects the frequency, causality and severity of DILI [8] for each drug, is challenging. Given the diversity of clinical manifestations of DILI, the uncertainty in causality assessment, severe underreporting of DILI cases, and the uncommon occurrence of DILI, it is difficult to identify a single resource that could provide all the information required for an accurate DILI annotation [8]. The approaches to annotate DILI risk are categorized as case report based, drug compendium based, or