Biological processes are complex, and that means that they lead to multi-dimensional data that human beings struggle to wrap their heads around. The good news is that AI is the perfect tool to spot patterns in this kind of data, and startups are increasingly using it to improve drug discovery from start to finish, from identifying biological targets to pre-clinical testing.
- Data mining millions of molecular structures
- Designing new molecules
- Predicting toxicity and off-target effects
- Predicting dosages of experimental drugs
- Mining molecular structures in silico
- Analyzing cellular assays in massive scale
- Faster development of more potent molecules
- AI chemical screening of the molecules
- Predicting chemical reactions via machine learning
- Sifting through chemical databases to find the right compounds
- Analyzing genomes and finding disease-causing variants
- Understanding the functions of non-protein coding sections
- Analyzing interactions between genes
- Determining the patterns of genome methylation
- Planning genome editing and CRISPR targets
- Assessing tumors and sequencing data from tumors
- Analyzing compounds to find a good fit for the targets
- Assessing compounds for the right dose
- In-silico labeling of cells and their structures for microscopic analysis
- Ghost cytometry: sorting cells through machine learning
- Assisting in image reconstruction for out of focus images or under-sampled data to amplify the data for interpretation
The AI drug discovery market is expected to be worth $40 billion by 2027. Discovery is the largest expenditure for drug development, so if we’re able to increase the quality of drug candidates, we’ll be able to improve the success rates of clinical trials. A number of well-funded companies have been making noise in this space.
Companies like Atomwise, Exscientia, InSilico Medicine, Insitro, HealX and Cyclica have been pursuing partnerships with big pharma along with the development of their own compounds. There are two main business models that we’re starting to see here. First, there’s the biotech startups that are using internal research and development to discover new drugs. Second, there’s the SaaS startups that are selling analytics software to pharma companies
In spite of all this, the progress so far has been underwhelming and partnerships between big pharma and these companies haven’t produced any breakthrough drugs. That’s not unexpected as these partnerships are still in their early stages. However, given the large amounts of capital invested so far and the expectations of investors to see results, the requirement for further large investments could be in jeopardy. In the next post, we can tackle which business models have shown good results in making progress toward better, faster, and cheaper drug discovery.
