|
Debora Marks
@
deboramarks
boston
|
|
Mission: develop AI for design of biological interventions for environment and human health (to save the world)
|
|
|
635
Tweetovi
|
270
Pratim
|
878
Osobe koje vas prate
|
| Tweetovi |
|
Debora Marks
@deboramarks
|
25. sij |
|
Ahh - it may start generating on its own ..
|
||
|
|
||
|
Debora Marks
@deboramarks
|
24. sij |
|
And how about biological sequence design? biorxiv.org/content/10.110…
|
||
|
|
||
|
Debora Marks
@deboramarks
|
24. sij |
|
How about the first and best for all protein mutation effects ?nature.com/articles/s4159…
|
||
|
|
||
|
Debora Marks
@deboramarks
|
24. sij |
|
Even has almost identical accessory genes to 2018 bat ‘SARS’ (not all in new database annotations)
|
||
|
|
||
| Debora Marks proslijedio/la je tweet | ||
|
Andy Biotech
@AndyBiotech
|
23. sij |
|
Important paper out on #Wuhan #nCoV2019 via @biorxivpreprint
96% sequence identity to a bat #coronavirus
~80% sequence identity to SARS
~95% AA sequence identity in 7 conserved replicase domains to SARS
uses the same cell entry receptor, ACE2, as SARS
biorxiv.org/content/10.110… pic.twitter.com/dGEjBvGx4j
|
||
|
|
||
|
Debora Marks
@deboramarks
|
24. sij |
|
PS: 99.99% proteins are outside CASP
|
||
|
|
||
|
Debora Marks
@deboramarks
|
24. sij |
|
#nCoV2019: does enalapril work?
|
||
|
|
||
|
Debora Marks
@deboramarks
|
24. sij |
|
#nCoV2019: Anyone notice two ‘accessory’ ORFs not annotated ? Orf13 (aka 9b) (28284:28577) and Orf14 (28734:28955) ~90% to BAT SARS 2017
|
||
|
|
||
|
Debora Marks
@deboramarks
|
19. sij |
|
That’s why we have to find new ways to infer homology or similarity - and new ways to find structural plasticity twitter.com/LindorffLarsen…
|
||
|
|
||
|
Debora Marks
@deboramarks
|
18. sij |
|
Predicting structure of designed protein is not harder than natural ones esp when designed have encoded the rules we know - take care with logic there folks ..
|
||
|
|
||
| Debora Marks proslijedio/la je tweet | ||
|
Lior Pachter
@lpachter
|
18. sij |
|
Diversity matters liorpachter.wordpress.com/2020/01/17/div… pic.twitter.com/FrufMlQ6xq
|
||
|
|
||
|
Debora Marks
@deboramarks
|
18. sij |
|
Ah .. that’s why I’ve always been such an unhappy person .. twitter.com/ribosaur/statu…
|
||
|
|
||
|
Debora Marks
@deboramarks
|
18. sij |
|
Thanks all who responded and also thanks to @DeepMindAI for apologizing. twitter.com/deboramarks/st…
|
||
|
|
||
|
Debora Marks
@deboramarks
|
17. sij |
|
|
||
|
Debora Marks
@deboramarks
|
17. sij |
|
|
||
|
Debora Marks
@deboramarks
|
17. sij |
|
Thx
|
||
|
|
||
|
Debora Marks
@deboramarks
|
17. sij |
|
Not sure it has systematically. We have done some internally and should publish
|
||
|
|
||
|
Debora Marks
@deboramarks
|
17. sij |
|
I’ll check that out - cool if not just indirect training test thing
|
||
|
|
||
| Debora Marks proslijedio/la je tweet | ||
|
Kresten Lindorff-Larsen
@LindorffLarsen
|
17. sij |
|
Thanks. I'm surprised that there is no analysis of the effect of the depth of the alignment and generalisability to proteins with fewer homologues. How much does the method learn something general and how much is based on a more robust extraction of dist distributions from MSAs?
|
||
|
|
||
|
Debora Marks
@deboramarks
|
17. sij |
|
Testable? #alphafold relies dependencies across evolution - simple. Only other way is real physics.
|
||
|
|
||