>>12080766>At least during manual design / synthesisis the synthetic steps are already understood and can be refined when upscaling manufacturing processes.Yes, this is actually a challenge our company is going to try and tackle in the next year, is rejecting any predicted compound that doesn't have a known/"easy" synthetic pathway already available.
>>12081958>computing ability and the bottle neck of actual synthesisBingo. Another big one is actually data curation, lots of data that just takes lots of man-hours to extract useable datasets out of for modeling.
>How much experience in the industry do you have?Not a lot, actually. I spend 2 years in drug discovery after my BS, went back to get my PhD (cell biology), split my work ~50/50 with machine learning, then just 6 months ago got hired to a startup. Ironically, even though I'm mostly experimental, I'm the best coder at the place (they have prebuilt software to make the machine learning extremely streamlined which is great, but it means the people hired just aren't great at actual code writing).
>What's your background/education?Wanted to do CS in college, switched halfway to biology (CS minor), went into drug discovery (mostly running assays, but also fitting dose-response curves), grad school for cell biology (focused on neuron development; mainly applying computation image analysis to developmental neuro), then internship at a startup, showed them I was good at shit, then got hired.
So probably less experience than most, but I wrote a solid portion (~80%) of one of the two grants big grants that are now funding the projects the startup is pursuing over the next few years, so I have a good grasp on capabilities.
