Thread by @Biotech2k1

Taking a look at the use of Tech in Biotech.

Here we will look at the use of Automation, Artificial Intelligence and Machine Learning in the Discovery of new drugs.

1/ Introduction:

It can take up to 10 years to develop a new drug. A majority of that time is spent in the lab testing targets and ideas. Its the process of sorting through thousands of potential targets to find the one that is worth advancing.

2/ It can cost a ton of money to develop a drug. The statistics show that it can cost upward of $2.5 billion to bring a new drug all the way to commercial. The high level of failure and high costs of testing and discovery is one space where technology can really help.

3/ It can improve the success rate. Computers can use databases and algorithms to scan thousands of potential targets to pick just a handful of targets that will have the highest chance of success. This can save a huge amount of time where trial and error would have been used.

4/ We are entering an era of new drug discovery that uses AI and ML to make it faster, more cost effective and more successful. The use of automation, AI, and ML can reduce the time and costs while improving the rate of success for new drug discovery.

5/ Its all about data:

Its all about collecting, organizing and using data to discover new targets

6/ - Genetic Data
Genetic data can predict what genes may drive a disease. It may also predict the variation of genes across the population and how they can affect that disease.

7/ Genetic data can help identify which patients will react best to specific treatments and which patients might have adverse reactions to specific treatments. Genetics and mutations play a major role in oncology and drug discovery.

8/ - Patient Data
Collecting patient data, prior therapies, side effects, and the outcomes can help predict which patients will respond best to which therapies and which can have bad outcomes. Combined with genetic data, we can personalize medicine for each patient.

9/ - Clinical Trial Data
Clinical data can be used by AI to determine what has worked or failed before. It can come up with new ways to potentially target the same target with a different approach. AI can also predict which combinations of drug might work best for each patient.

10/ - Scientific Data
This can help direct AI where to look for new targets. You might have data where something worked, but it was too toxic or had off target effects. The AI might be able to use the combination of all the data to predict a new target.

11/ Uses of AI and ML:

The data can take tens of thousands of potential targets and screen them down to a handful of best potential candidates. This can save a lot of time and cost from using the old methods of trying each potential target to get the one that works.

12/ Even then, they often did not get the best possible candidate. With AI, companies can spend more time with screening targets before they ever move into the lab where its cost significantly more.

13/ Computers can be used with software to build models used in drug discovery. There are companies like $SDGR and $RLAY using modeling software to predict how proteins and enzymes behave and move.

14/ They use this to design new drugs that target the mutated proteins that drive cancer growth. They can develop targeted therapies that take on these genetic mutations in cancer.

15/ Companies like $EXAI are using ML to take patient genetic information and develop better combinations of already existing drugs to improve outcomes. They showed that using ML could improve the outcomes of patients by up to 30%.

16/ They also use AI to help screen genetic data to find valid targets for new therapies.

17/ Another use of modeling in drug discovery is predicting drug metabolism. There are certain drugs that can be toxic when metabolized. Knowing which ones will have toxicity use to be done by trial and error.

18/ Now there is software that can predict which compounds might end up being toxic. This is another example where AI can help reduce a lot of the failure and cost in drug discovery.

19/ Automation:

Automation of the Lab brings many benefits from increasing scale of testing, reducing costs, saving time and reducing error rates.

20/ Companies like $BLI are developing machines that can automate many of the tedious and time consuming tasks in the lab. They build machines around sorting and selecting cells.

21/ These machines can make a hybridoma for monoclonal antibodies at 10x the scale and with far higher success in 90% less time. This is an example of how automation is changing drug discovery.

22/ Companies like $RXRX are using robots to conduct millions of experiments per week and recording all the data into a super computer.

23/ That super computer can then use AI to predict drugs targets for clinical development. It can screen known chemical entities for new uses or even predict new targets to be explored.

24/ Conclusion:

Using automation, AI and ML can allow companies to build a platform of technology that can lower costs and improve success. It enables companies to do more with less. They can build a lab with a few highly skill individuals that can drive innovation.