BEGIN:VCALENDAR
VERSION:2.0
PRODID:www.dresden-science-calendar.de
METHOD:PUBLISH
CALSCALE:GREGORIAN
X-MICROSOFT-CALSCALE:GREGORIAN
X-WR-TIMEZONE:Europe/Berlin
BEGIN:VTIMEZONE
TZID:Europe/Berlin
X-LIC-LOCATION:Europe/Berlin
BEGIN:DAYLIGHT
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
DTSTART:19810329T030000
RRULE:FREQ=YEARLY;INTERVAL=1;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
DTSTART:19961027T030000
RRULE:FREQ=YEARLY;INTERVAL=1;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:DSC-21624
DTSTART;TZID=Europe/Berlin:20250605T130000
SEQUENCE:1749101935
TRANSP:OPAQUE
DTEND;TZID=Europe/Berlin:20250605T140000
URL:https://www.dresden-science-calendar.de/calendar/de/detail/21624
LOCATION:TUD CRTD\, Fetscherstraße 10501307 Dresden
SUMMARY:CMCB Life Sciences Seminar: Prof. Anne-Florence Bitbol\, EPFL\, Ins
tiute of Bioengineering\, Switzerland
CLASS:PUBLIC
DESCRIPTION:Speaker: \nInstitute of Speaker: \nTopics:\nBiologie\, Willkomm
en\n Location:\n Name: TUD CRTD ()\n Street: Fetscherstraße 105\n City
: 01307 Dresden\n Phone: +49 (0)351 458 82052\n Fax: +49 (0)351 458 8205
9 \nDescription: Host: James Sáenz
Tit
le: \"Predicting interaction partners and generating new protein
sequences using protein language models\"
Abstract: Protein language models trained on multiple sequence alignments of homol
ogous proteins successfully capture coevolution between amino acids in str
uctural contact: this is one of the ingredients of the success of AlphaFol
d. We have used such models\, especially MSA Transformer\, to generate new
protein sequences from given protein families\, and to predict which prot
eins interact among the members of two protein families. Despite their suc
cesses\, a drawback of models based on multiple sequence alignments is tha
t sequence alignment can be imperfect. Thus\, we are developing ProtMamba\
, a homology-aware but alignment-free protein language model. ProtMamba st
arts from concatenated homologous sequences\, is based on the Mamba state-
space model architecture\, has promising generative properties\, and is ab
le to predict fitness.
DTSTAMP:20260710T173729Z
CREATED:20250110T063913Z
LAST-MODIFIED:20250605T053855Z
END:VEVENT
END:VCALENDAR