L5: AI in biotechnology

Question 1

technologies alone can transform healthcare. T/F?

Answer 1

False

Question 2

challenges of innovation [4]

Answer 2

1. high cost
2. long development time
3. low success rate
4. difficult to determine which drug and dose to use

Question 3

between development and discovery, which stage takes up a longer time? state what is done during this stage.

Answer 3

discovery.

identify drug target, generate leads, optimise them

Question 3

how can AI help in the discovery of new drugs?

Answer 3

1. aid in drug discovery: target validation and lead optimisation
2. drug development during pre-clinical and clinical: help find out the best combinations of drugs and doses [optimisation]
3. N of 1: personalised therapy: optimise the duration of treatment and drugs according to patient
4. digital medicine: combat cognitive decline by having programs for patients

Question 3

what is N-of-1 Healthcare?

Answer 3

to use a patient’s own data to manage their own treatment

Question 3

identifying synergy alone is enough for optimisation. t/f?

Answer 3

false. intra-patient drug synergism and antagonism is dose dependent

Question 3

why do clinical trials fail so often?

Answer 3

drug synergy and antagonism needs to be re-optimised AT EACH STAGE OF DEVELOPMENT to de-risk drug development

Question 3

how can we use AI to globally optimise drug development?

Answer 3

1. finding the optimum combination of drugs from a massive drug-dose space
2. drug interaction analysis: can find unforeseen interactions for evolving drug sets

Question 4

differentiate between traditional and current small data AI-based optimisation

Answer 4

Traditional:
- based on pre-existing data
- FIXED dose
- find the best option out of n number of combinations

Small data:
- less data points [still have diff combinations]
- but have variation in dose –> find best combination out of n no. of assays representing 1 trillion++ possibilities
- use actual expt results, not pre-existing data

Question 5

Perks/use of using Identif.AI

Answer 5

1. find best permutation of drugs and dose
2. ranks the combination–> know which combi to avoid
3. dynamic analysis wrt to patient’s own progress and reaction to drug

Question 6

how can AI make healthcare more efficient and affordable? [4]

Answer 6

1. guide treatment choices -> personalise medicine for patients
2. more efficient diagnosis [lesser time and cost]
3. improve clinical trials optimisation and drug development
4. empowering patient