KBase Science Session: Data integration to support (or refute) predictions скачать в хорошем качестве

KBase Science Session: Data integration to support (or refute) predictions

Much of biological science knowledge is based on prediction – algorithmic assessment of taxonomy and function, determined by parameters based on current knowledge. But how certain are we that those parameters are accurate across the tree of life, across ecosystems, and across time. By exploring and integrating multiple lines of evidence (i.e., multi-omics data, experimental validation, mechanistic modeling, machine learning, etc), we can build a story that supports or refutes our predictions. This session will feature examples of research questions and tools integrating data with the goal of building trust in our assertion of the world around us. Speakers: 00:00 Session Introduction: Elisha Wood-Charlson (Lawrence Berkeley National Laboratory) Why bother with data integration, and why does my data matter? 03:52 Dale Pelletier (Oak Ridge National Laboratory) Measuring microbial phenotypes for improving genome-based predictions 21:28 Shinjae Yoo (Brookhaven National Laboratory) Knowledge extraction from literature 35:39 Chris Henry (Argonne National Laboratory) Predicting protein function using structure and sequence similarity in KBase 51:10 Bill Nelson (Pacific Northwest National Laboratory) Integrating data to predict functions for gaps in metabolic models 01:08:10 Paramvir Dehal (Lawrence Berkeley National Laboratory) Leveraging LLMs to synthesize and develop new questions 01:30:01 Elisha Wood-Charlson (Lawrence Berkeley National Laboratory) Getting credit for contributions in a big data world