Mutational effect transfer learning (METL) is a protein language model framework that unites machine learning and biophysical modeling. Transformer-based neural networks are pretrained on biophysical Show more
Mutational effect transfer learning (METL) is a protein language model framework that unites machine learning and biophysical modeling. Transformer-based neural networks are pretrained on biophysical simulation data to capture fundamental relationships between protein sequence, structure and energetics. Show less
Abstract Imaging contrast agents are widely investigated in preclinical and clinical studies, among which biogenic imaging contrast agents (BICAs) are developing rapidly and playing an increasingly i Show more
Abstract Imaging contrast agents are widely investigated in preclinical and clinical studies, among which biogenic imaging contrast agents (BICAs) are developing rapidly and playing an increasingly important role in biomedical research ranging from subcellular level to individual level. The unique properties of BICAs, including expression by cells as reporters and specific genetic modification, facilitate various in vitro and in vivo studies, such as quantification of gene expression, observation of protein interactions, visualization of cellular proliferation, monitoring of metabolism, and detection of dysfunctions. Furthermore, in human body, BICAs are remarkably helpful for disease diagnosis when the dysregulation of these agents occurs and can be detected through imaging techniques. There are various BICAs matched with a set of imaging techniques, including fluorescent proteins for fluorescence imaging, gas vesicles for ultrasound imaging, and ferritin for magnetic resonance imaging. In addition, bimodal and multimodal imaging can be realized through combining the functions of different BICAs, which helps overcome the limitations of monomodal imaging. In this review, the focus is on the properties, mechanisms, applications, and future directions of BICAs. Show less
Preclinical strategies that are used to identify potential drug candidates include target-based screening, phenotypic screening, modification of natural substances and biologic-based approaches. To in Show more
Preclinical strategies that are used to identify potential drug candidates include target-based screening, phenotypic screening, modification of natural substances and biologic-based approaches. To investigate whether some strategies have been more successful than others in the discovery of new drugs, we analysed the discovery strategies and the molecular mechanism of action (MMOA) for new molecular entities and new biologics that were approved by the US Food and Drug Administration between 1999 and 2008. Out of the 259 agents that were approved, 75 were first-in-class drugs with new MMOAs, and out of these, 50 (67%) were small molecules and 25 (33%) were biologics. The results also show that the contribution of phenotypic screening to the discovery of first-in-class small-molecule drugs exceeded that of target-based approaches - with 28 and 17 of these drugs coming from the two approaches, respectively - in an era in which the major focus was on target-based approaches. We postulate that a target-centric approach for first-in-class drugs, without consideration of an optimal MMOA, may contribute to the current high attrition rates and low productivity in pharmaceutical research and development. Show less