Akanksha Yadav @akankshay58@qoto.org
- https://twitter.com/akankshay58
- Website
- https://akankshay.github.io/
grad student at Caltech Bioengineering | prev research tech at HHMI Janelia | chemistry, systems biology, protein engineering, athlete, episodic poet | MSc Molecular biology, MPI-NAT l BS Chemistry, IIT Bombay https://twitter.com/akankshay58
Joined Oct 2022
Helpful paper unpacking some metrics and intuition behind higher success rate (10-fold?) of designing protein binders using DL and/or physics-based methods alone
Improving de novo protein binder design with deep learning | Nature Communications https://www.nature.com/articles/s41467-023-38328-5
Neat single-molecule experiments!
Two-component molecular motor driven by a GTPase cycle | Nature Physics
https://www.nature.com/articles/s41567-023-02009-3
---
RT @stpalli
The mechanism of this new motor is distinct from conventional ones.
Unlike motors such as Kinesin, it runs on GTP not ATP.
Also, the motor has no "power stroke" motion of one protein domain relative to another, but relies on a sudden change in the stiffness of the en…
https://twitter.com/stpalli/status/1654141624933335048
Super cool observations! Not entry, not copy #, but lack of histone modifications affecting transgene expression
The AAV capsid can influence the epigenetic marking of rAAV delivered episomal genomes in a species dependent manner | Nature Communications https://www.nature.com/articles/s41467-023-38106-3
Fun read! Emergence of Cooperative Glucose-Binding Networks in Adaptive Peptide Systems https://pubs.acs.org/doi/10.1021/jacs.3c01620#.ZEXPMrgBLHs.twitter
In vivo bioluminescence imaging of natural bacteria within deep tissues via ATP-binding cassette sugar transporter | Nature Communications https://www.nature.com/articles/s41467-023-37827-9
Genetically encoded barcodes for correlative volume electron microscopy | Nature Biotechnology https://www.nature.com/articles/s41587-023-01713-y
Neoepitope formation through the generation of RNA-derived “editopes” https://www.biorxiv.org/content/10.1101/2023.03.16.532918v1
Thanks so much Tim @m8ta
I want to especially thank my BB405,503 UG course professors @patankar_lab @Banerjeeiitb, who taught molecular biology in a way that probably changed the course of my academic interests ever since.
Also excited to get 'The Eighth Day of Creation' by Judson as my celebratory book
It has been a tortuous path to get here, but I feel very fortunate - elated to share that I will be pursuing my Ph.D. in Bioengineering at @Caltech starting this fall. I am thankful to numerous people including my mentors, who gave me a chance to work with them amidst pandemic
RT @MuirLab
Looking to study chromatin interactomes with proximity labeling? Check out our new paper in @Nature, where we combine our split-intein expertise with uMap technology from the @MacMillan_Lab to map the microenvironment of chromatin state changes.
A Vaccinia-based system for directed evolution of GPCRs in mammalian cells | Nature Communications https://www.nature.com/articles/s41467-023-37191-8
Tried tracing trends: concept, molecules, methods
(via word search on PubMed)
RT @shivam_kajale
Delighted to share that I recently had the chance to talk about my work on development of highly energy-efficient “beyond CMOS” devices, based on magnetics, spintronics and magnetoelectrics, at @TEDxBoston 's Planetary Stewardship forum.
Akanksha Yadav
boosted
For writing #tootorial, I’ve used the trick where the first toot is Public but the rest are Unlisted. The top post will be the one that shows up in followers timelines, rather than all of them in reverse order.
I think the Unlisted ones can be boosted by others, but they aren’t searchable though, which is a downside.
Papers and patents are becoming less disruptive over time | Nature https://www.nature.com/articles/s41586-022-05543-x
Akanksha Yadav
boosted
@ypriverol had a conversation with a new data science student looking at a high throughput screen. They asked, how do you approach this as a statistician? I said, make plots, more plots, keep making plots. Raw data, summary stats, typical features/cases, exceptional features/cases. Student asked, can AI help in this process? Me: not really, this is where the human 👁️ and 🧠 are your best tools, to then provide quality data for models to learn meaningful patterns instead of artifacts and bias
For eg, in the last few days,
{I have been listening to talks on BCI, explaining to friends about an experience when I wanted to understand someone’s approach to a problem, thought about watching one of the Harry Potter movies, read about José’s blog embeddings}
..hence,Pensieve?
The key point being that the dissection is in a way that it is immediately obvious to understand your idea/thought/feeling and how you arrived at it at the moment, if you/anyone revisited it in future, and in a way that cannot be corrupted by any biases/assumptions
I wonder what a Pensieve* looks like for different people? Do people dissect their ideas, referencing it all the way back to its ‘origin’? Do people record this process explicitly?
(*dish from Harry Potter in which one could review any extracted memories from a third-person pov)
- https://twitter.com/akankshay58
- Website
- https://akankshay.github.io/
grad student at Caltech Bioengineering | prev research tech at HHMI Janelia | chemistry, systems biology, protein engineering, athlete, episodic poet | MSc Molecular biology, MPI-NAT l BS Chemistry, IIT Bombay https://twitter.com/akankshay58
Joined Oct 2022
