DailyPost 2970
ALPHAFOLD3 GOES OPEN SOURCE
There were limits to the usage and the final utility of open source. At times it needed either customisation or support, which came with its own complications. Making open source work for you has its own tool kit. Leaving aside the intricacies of the open source world for the time being, what is to be celebrated is that AlphaFold3 has gone open source. Besides knowing of the granular details and the fine print, given its progeny, I would believe that it is a ready to use scientific model, set to revolutionise the area of drug discovery.
AlphaFold3 is the winner of the 2024 Chemistry Nobel Prize shared by John Jumper, who leads AlphaFold and DeepMind CEO Demis Hassabis. What is AlphaFold3? It is the code underlying the Nobel-prize winning tool, for modelling protein structures which can now be downloaded by scientists. A tool which is the product of an out of the world AI research being available for free, to further research and consequently drug discovery is a boon for mankind. “We’re excited to see what people do with this,” an excited John Jumper said.
The current AlphaFold3 open source release waives off all the terms and conditions of the past for the scientists to use its full potential. How much has the AlphaFold3 world changed in just six months? Might be the Nobel-prize was an accentuating factor. Google DeepMind had controversially withheld code from a paper describing the protein structure prediction model only six months back. “AlphaFold3, unlike its predecessors, is capable of modelling proteins in concert with other molecules.” In the first instance unlike in AlphaFold2, where the underlying code was released, in the case of AlphaFold3 earlier access via web server was provided “that restricted the number and types of predictions scientists could make”
The AlphaFold3 server in its operations prevented scientists from predicting “how proteins behave in the presence of potential drugs.” Unfortunately, this happens to be at the core of this revolutionary AI tool. In this backdrop DeepMind’s decision can really be termed as a watershed moment in the direct use of AI for mankind. The impact of the decision to release code now means academic scientists can predict such interactions by running the model themselves. Isomorphic Labs, a DeepMind company, has been applying AlphaFold to drug discovery. The aim is that anyone would be able to download the AlphaFold3 software code for non-commercial use. For now, “only scientists with an academic affiliation can access the training weight on requests.”
ALPHAFOLD3 THUS EPITOMISES THE PHILOSOPHY OF ARTIFICIAL INTELLIGENCE (AI) FOR HUMANITY.
Sanjay Sahay
Have a nice evening.