←
Michael Peer258 posts
Michael Peer avatar

Michael Peer

@michael_peer_

ML & Computational Biology Lead @ Biolojic Design: using AI to design antibody therapeutics. Previously a neuroscientist researching brain networks and memory.

◷ Joined August 2016▧ 33 Photos and videos
Michael Peer avatar
Michael Peer@michael_peer_·Sep 8

On a personal note, I am proud to be a part of this effort by the whole Biolojic Design team. Multibodies can be a game-changer to medical treatments across a variety of clinical fields, and I am looking forward to seeing them improve patients' lives soon.

Post attachment
Michael Peer avatar
Michael Peer@michael_peer_·Sep 8

The AI challenge was significant - achieving precise binding to two targets, on a small antibody surface that can usually only bind one. The solution required multiple innovations across data collection and modeling and a joint effort between our computational and wet lab teams.

Michael Peer avatar
Michael Peer@michael_peer_·Sep 8

Our multibody AI platform is highly scalable - we managed to successfully design multibodies against eight different target pairs, involved in different cancer and autoimmune disorders. Two of our multibodies have already advanced and will soon enter clinical trials in patients.

Michael Peer avatar
Michael Peer@michael_peer_·Sep 8

We have used AI to develop a new antibody format that is extremely rare in nature - multibodies, where each arm can bind two different targets. Multibodies keep the stable, long-lasting and bivalent format of a regular antibody, but with the ability to bind two different targets.

Post attachment
Michael Peer avatar
Michael Peer@michael_peer_·Sep 8

To address this problem, companies have started to combine antibody fragments to target several pathways simultaneously. This opens new and exciting therapeutic capabilities, but reduces the stability, durability, bivalency, and manufacturing ease compared to standard antibodies.

Post attachment
Michael Peer avatar
Michael Peer@michael_peer_·Sep 8

Antibodies are among the world's most successful therapeutics: they are stable, long-acting, and enable precise targeting of almost any biological pathway in the body. However, each antibody is specific to only one target, thus limiting their potential to treat complex disorders.

Michael Peer avatar
Michael Peer@michael_peer_·8 Oct 2023

I usually don’t write political posts on Twitter, but I wanted to clarify a few things about the situation in Israel.

Post attachment
Michael Peer avatar
Michael Peer@michael_peer_·17 Sep 2023

An impressive new work by the great Seth Koslov on medial parietal cortex subdivisions!

Ever notice medial parietal cortex in neuroimaging studies of episodic memory and wonder what's going on there? In our preprint (biorxiv.org/content/10.1101/…) we identify interleaving subregions responsive to memory-decisions vs semantic content during recognition within individuals.

The Epstein Lab avatar
⟳ Michael Peer retweeted
The Epstein Lab@epstein_lab·5 Apr 2023

Come join us in Philly! @epstein_lab @Penn has postdoc openings on NIH-funded projects that use advanced fMRI methods to investigate the neural basis of cognitive maps (broadly defined), spatial navigation, and spatial/nonspatial memory in humans. Collaboration with @NoraNewcombe

Thomas Yeo avatar
⟳ Michael Peer retweeted
Thomas Yeo@bttyeo·17 May 2022

1/9 Paper is out rdcu.be/cNHoL Here's some takeaways including new tidbits (good & bad!) since the preprint & even after the paper was accepted. Large sample sizes are helpful for individual-level prediction. We have known that for quite some time, but the recent...

1/5 Our new preprint doi.org/10.1101/2020.08.10.2… uses meta-learning to boost functional connectivity based prediction of *NEW* phenotypes in small samples. We achieve 100%-400% improvements in many scenarios! @hetong93 @AvramHolmes @anlijuncn @danilobzdok @AvramHolmes @INM7_ISN

The Epstein Lab avatar
⟳ Michael Peer retweeted
The Epstein Lab@epstein_lab·14 Jun 2022

We are recruiting participants for fMRI experiments! If you are in the @Penn / Philly area and want to earn up to $95, you can find more info and register here: 🔗forms.gle/d1QmsS6Vj9vXQ78Q6