Reinventing the wheel with web3, drug research edition

Bybit
Reinventing the wheel with web3, drug research edition
BTCC


Martin Shkreli is back to his entrepreneurial ways, just a couple of months after his release from jail.

Shkreli announced on Monday that he’s launching a web3 platform called Druglike. This is not a drug company, mind you – he’s banned from the pharmaceutical industry – but an open-source platform for drug research, according to a press release:

Druglike is building a decentralised computing network which provides resources for anyone looking to start or contribute to early-stage drug discovery projects. Unlike competitors, Druglike will be web-based and completely free to use. Druglike will soon release a web-based suite for target identification, drug design, and tools for both constructing and running large-scale virtual screening workflows.

Open-source medical research platforms already exist, of course. 

For example, one platform called VirtualFlow was used for a 2020 paper in Nature, where a team of academics “prepared one of the largest and freely available ready-to-dock ligand libraries, with more than 1.4bn commercially available molecules.” A quick (and we hope not flawed) survey of the internet helps us guess that ligands bind to proteins in order to target particular pharmaceutical therapies, and docking is a modelling technique to predict how ligands and proteins will bind. 

Its white paper references the Nature experiment in the bolded section below:

Other screens of similar size run on cloud-provisioned hardware, often using tens of thousands of CPU cores for multiple days at a time… On-premises supercomputers are also used for running large-scale virtual screening pipelines… These large-scale screening projects require either special discounted deals with cloud compute providers, or prioritized academic access to high performance computing (HPC) resources. Smaller research groups may run screens on commodity cloud computing resources, but the profit margins on specialized, cloud-provisioned hardware usually constrain the size of the screen.

Druglike uses ligand docking as the example in its white paper. What sets it apart from VirtualFlow, from what we can tell, is that they add a heap of web3 to the mix, along with some decentralised computing. The whitepaper describes ligand-docking modelling as an optimisation problem, and says the project creates a new protocol called “Proof-of-Optimisation” – basically proof-of-work for function-optimisation problems – for verification. 

We will confess we don’t currently have the data-science chops to independently evaluate how realistic this proposal is. 

But when it comes to decentralised computing, we did catch a whitepaper reference to Folding@home, a decentralised project where regular people can donate computing power to help university labs run protein-folding simulations. 

Folding@home teams offer different “rewards” for participation, but there doesn’t appear to be a standard for what rewards are used. In other words, there could realistically be zero cryptocurrency involvement at all. In early 2019, Folding@home published a cryptocurrency policy that says that outright: 

In light of recent discussions within the Folding@home community concerning our relationship to cryptocurrencies, we would like to take this opportunity to make an official statement. Folding@home does not endorse any cryptocurrencies at this time but we are happy to engage with participants with a wide variety of motivations. Whether you’re interested in curing diseases, advancing our understanding of how the world works, seeing who can build the fastest computer, or earning cryptocurrency. There’s a long history of participants in Folding@home developing third party applications that enhance the community, and we see cryptocurrencies as falling in this category. Our policies will always be dictated by the science that is the foundation of Folding@home, but maintaining a level of consistency that incentivizes everyone to continue participating in the project is a high priority.

On forums for new and interested volunteers, one post from October 2021 comes from a user saying that they heard “this thing doubles as a dogecoin miner,” and asking “how do I get doge out of FaH, and how do I join a team?”

We also can’t help but notice that Druglike’s press release is followed by a “safe harbour” statement, just in case the tokens planned by the platform are deemed to be securities. Shkreli is also banned from serving as an officer or director of public companies, to be sure, so the whitepaper includes a disclaimer that says “if and when Druglike makes . . . an offering in the United States, the offering likely will be available solely to accredited investors.”



Source link

fiverr
fiverr

Be the first to comment

Leave a Reply

Your email address will not be published.


*