Presentation
KBase Research and Genome Annotation Agent
Presenter
DescriptionWe leverage the latest developments in Large Language Models (LLMs) to create a KBase Research Agent, guiding users in navigating and analyzing data within the KBase platform. This Agent currently converses with users, interprets their data and its relationship with public data on the KBase system, and uses this information to help users reach their analysis goals. It does this by building a custom analysis plan by discussing options with users using the human in the loop tool and orchestrates the execution of the plan through the KBase jupyter notebook-like Narrative interface. Using KBase apps, the agent will help assemble and annotate the reads, ensuring quality control at each step. It will interpret outputs and guide each subsequent step in the workflow. This agent enables LLM-driven acceleration in making a large number of private fully sequenced microbial genomes to become publicly available, significantly lowering the barrier.
TimeTuesday, June 1710:30 - 11:00 CEST
LocationCampussaal - Plenary Room
Event Type
PASC Poster
Authors
