Toxicology & Lead Generation Flashcards
What are the 4 ways to discover a new drug?
- Follow-on compounds
- computational modelling and sequencing
- serendipitous discovery
- evergreening
What are follow-on compounds?
- new indication discovered
- need to get lucky for this to happen
What is computational modelling and sequencing?
- increasing number of solved protein x-ray crystal structures
What is serendepitous discovery?
- same mechanism as previously reported drug
- “fast followers” or “me too” drugs
- provide professionals and patients with options, keep prices low and comes with a degree of reassurance
What is evergreening?
- extreme form of a “me too” drug
- requires an in depth knowledge of medicinal discovery
- extending the duration of a patent with minimal chemical intervation
What is target validation?
- first step
- exploring relationship between pharmacological modulation of a target and a pathological condition
- what are you trying to achieve?
What is hit identification?
- second step
- finding a chemically accessible (able to be synthesised easily) compound displaying an initial activity towards a target
What characteristics does a compound have for hit identification?
- small organic molecule
- moderate affinity
- low MW (150 < 400)
- c Log P < 4.5
- 1-4 rings
- < 8 H bond acceptors
- < 5 H bond donors
What is rational design?
- based on physiological binders (substrates, co-factors) gives us idea of shape as it binds to target of interest already
- utilises structural info to improve ligand interactions in binding site
- smaller less complex molecules with a smaller number of functional groups and tend to rule of 3
What is the rule of 3?
- MW < 300
- cLogP <= 3
- HBDs <= 3
- HBAs <= 3
What is high throughput screening?
- if you don’t know much about your target rational design isnt useful
- screen as many compounds and hope
- high number of sp2 centres = general flat nature
What are the two types of high-throughput screening?
- unselected screens
- hit rates ~ 1%
- screen millions of compounds to get a decent number of hit families to follow up
- limited by budget, time, rescources and intrinsic throughput
- limited to a number of compounds at a single concentration
- generates noise (false positives or negatives)
- selected screens
- enough info about target to inform screening
- combination of rational design and unselected screening
- greatly increases speed, reduces costs and makes identification easier
How can you recognise natural compounds?
- high number of stereocentres and complex structure
- lots of fused-ring systems
- macrocycles
- presence of a number of basic Nitrogen atoms (alkaloids)
- chemical produced by organism
- trying to control synthesis is difficult
What are PAINs?
pan-assay interference compounds (PAINs)
- positive hit compounds which turned out to be due non-specific binding
- defined structures, containing several chemical classes
- time and research money wasted
- known PAINs bearing known troublesome chemical groups which have non-specific binding/false positives (eg quinones) are filtered out before screening
Why shouldnt you excludes PAINs completely?
structures can be important in final molecules
What is the rule of 3 derived from?
Lipinski’s rule of 5
What is fragment screening?
method of reducing ligand complexity to increase rhe chance of a match with target site
samples chemical space at finer resolution
What are prerequisites before starting a hit-2-lead campaign?
- chemically acceptable
- can you do this chemistry on scale?
- starting hit responds to chemical modulation, quantifiable SAR
- freedom to operate
- preliminary ADME profile looks favourable