BEGIN:VCALENDAR VERSION:2.0 METHOD:PUBLISH PRODID:-//Missouri State University/Calendar of Events//EN CALSCALE:GREGORIAN X-WR-TIMEZONE:America/Chicago BEGIN:VTIMEZONE TZID:America/Chicago BEGIN:DAYLIGHT TZOFFSETFROM:-0600 TZOFFSETTO:-0500 DTSTART:20070311T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU TZNAME:CDT END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:-0500 TZOFFSETTO:-0600 DTSTART:20071104T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU TZNAME:CST END:STANDARD END:VTIMEZONE BEGIN:VEVENT UID:dcf3808f-e2da-454d-a6c6-508cbe6b607e.229509@calendar.missouristate.edu CREATED:20230918T133208Z LAST-MODIFIED:20230918T133208Z LOCATION: SUMMARY:Computer Science Seminar - Applications of Prompt-Based Learning f or Biomedical Data Analysis DESCRIPTION:AI foundation models\, trained on large-scale data\, offer unp recedented opportunities for a wide range of applications. The potentials of these models are further magnified when combined with prompt-based le arning\, achieving state-of-the-art (SOTA) results even with minimal labe led data. \n\n\nThis presentation delves into the biomedical utilization of three such models: ChatGPT\, Segment Anything Model (SAM)\, and Evolut ionary Scale Modeling (ESM). Specifically\, we utilized SAM to identify k ey attributes of pathway entities and their relationships from the figure s of research papers. We queried ChatGPT to identify gene relationships b y designing effective prompts. To enhance the accuracy of ChatGPT's feedb ack\, we introduced an innovative iterative prompt refinement technique. This method assesses prompt efficacy using metrics like F1 score\, precis ion\, and recall. Based on these evaluations\, ChatGPT was re-engaged to suggest improved prompts. We also integrated prompt-based learning with S AM to detect proteins in cryo-Electron Microscopy (cryo-EM) images. Furth ermore\, we prompted the ESM model for signal peptide predictions and lar ge models of single-cell RNA-seq data for data analyses. The outcomes of our studies underscore the potential utilities of foundation models and p rompt-based learning for efficient biomedical data analyses and predictio ns.\n\n\nPresented by Dr. Dong Xu\, Curators' Distinguished Professor in the Department of Electrical Engineering and Computer Science at the Univ ersity of Missouri-Columbia.\n\n\nJoin Zoom Meeting https://missouristate .zoom.us/j/96796284909?from=addon Meeting ID: 967 9628 4909 X-ALT-DESC;FMTTYPE=text/html:<html><head><title></title></head><body><p><s pan>AI foundation models\, trained on large-scale data\, offer unpreceden ted opportunities for a wide range of applications. The potentials of the se models are further magnified when combined with prompt-based learning\ , achieving state-of-the-art (SOTA) results even with minimal labeled dat a. </span></p>\n<p><span>This presentation delves into the biomedical uti lization of three such models: ChatGPT\, Segment Anything Model (SAM)\, a nd Evolutionary Scale Modeling (ESM). Specifically\, we utilized SAM to i dentify key attributes of pathway entities and their relationships from t he figures of research papers. We queried ChatGPT to identify gene relati onships by designing effective prompts. To enhance the accuracy of ChatGP T's feedback\, we introduced an innovative iterative prompt refinement te chnique. This method assesses prompt efficacy using metrics like F1 score \, precision\, and recall. Based on these evaluations\, ChatGPT was re-en gaged to suggest improved prompts. We also integrated prompt-based learni ng with SAM to detect proteins in cryo-Electron Microscopy (cryo-EM) imag es. Furthermore\, we prompted the ESM model for signal peptide prediction s and large models of single-cell RNA-seq data for data analyses. The out comes of our studies underscore the potential utilities of foundation mod els and prompt-based learning for efficient biomedical data analyses and predictions.</span></p>\n<p><span>Presented by Dr. Dong Xu\, Curators' Di stinguished Professor in the Department of Electrical Engineering and Com puter Science at the University of Missouri-Columbia.</span></p>\n<p><spa n>Join Zoom Meeting<br> <a href="https://missouristate.zoom.us/j/96796284 909?from=addon">https://missouristate.zoom.us/j/96796284909?from=addon</a ><br> <br> Meeting ID: 967 9628 4909</span><span></span></p></body></html> DTSTART;TZID=America/Chicago:20230925T154000 DTEND;TZID=America/Chicago:20230925T164000 SEQUENCE:0 URL: CATEGORIES:Current Students,Faculty,Staff END:VEVENT END:VCALENDAR