MCHM3001 · From Molecules to Therapeutics
Macrocycles & Massively-Parallel Screening
Lecture 10 of MCHM3001 covers cyclic-peptide and macrocyclic drugs and the display technologies that screen enormous libraries at once. The examinable core is the combinatorial-diversity relationship (nˣ), how display platforms differ in achievable diversity and hit affinity, and why cyclic peptides make useful protein-protein-interaction inhibitors. It builds on the screening chapter and appears in Test 1/2 and the final.
What this chapter covers
- 01Massively parallel (display) screening: 10⁶ to ≥10¹² molecules at once, with DNA barcoding for identification
- 02Combinatorial mathematics: number of products = nˣ (n building blocks, x coupling steps)
- 03Cyclic peptides (romidepsin, cyclosporine) as drug-like PPI inhibitors with unusual chemistries
- 04Phage display of cyclic peptides (disulfide/dithioether/bicyclic linkers): library ~10⁹, KD/IC50 ~50–1000 nM
- 05mRNA display / RaPID: puromycin links peptide to mRNA; diversity limited only by concentration (>10¹⁴), KD/IC50 ~1–50 nM
- 06Genetic-code reprogramming for non-canonical residues
- 07Worked hits: HGF inhibitor HiP-8 (IC50 0.9 nM without optimisation), KDM4A inhibitor (IC50 42 nM)
- 08Trade-offs: cyclic-peptide display is cheap and drug-like in affinity but often not orally available or permeable
Sizing a split-pool library with nˣ
- +1Identify n and x: n = 20 building blocks (amino acids) and x = 3 steps (positions).
- +1Apply the combinatorial rule, products = nˣ: 20³ = 20 × 20 × 20 = 8,000 distinct compounds, one per bead.
- +1Note the identification problem: assaying 8,000 individual compounds is far more work than the synthesis, which is why split-pool needs a deconvolution strategy rather than brute-force testing.
- +1Explain deconvolution: instead of testing all 8,000, you test pools defined one position at a time and resynthesise the active sub-pools, so the active building block at each of the 3 positions is found in roughly 3 × 20 = 60 informative assays rather than 8,000 — recursion trades a little resynthesis for a huge cut in screening.
Key terms
- Combinatorial library size (nˣ)
- The number of distinct products from combining n building blocks over x steps; e.g. 20 amino acids over 3 positions gives 20³ = 8,000 compounds.
- Split-and-pool (DCR)
- Divide-couple-recombine solid-phase synthesis that generates all nˣ combinations while keeping one compound per bead; efficient to make but requiring deconvolution or encoding to identify hits.
- Massively parallel (display) screening
- Screening 10⁶ to ≥10¹² molecules simultaneously with a DNA barcode tying each molecule to its identity — the basis of phage and mRNA display.
- mRNA display (RaPID)
- A display method where puromycin covalently links a peptide to its own mRNA; because diversity is limited only by concentration, libraries can exceed 10¹⁴ and yield drug-like (1–50 nM) cyclic-peptide hits.
- Cyclic peptide
- A macrocyclic peptide (e.g. romidepsin, cyclosporine) with drug-like properties for protein-protein-interaction targets, though often lacking oral availability and passive permeability.
- Deconvolution
- The process of identifying which compound in a pooled library is active — by positional/recursive resynthesis of sub-pools or by chemical/non-chemical encoding tags.
Macrocycles & Massively-Parallel Screening FAQ
How do you calculate the size of a combinatorial library?
Use products = nˣ, where n is the number of building blocks available at each step and x is the number of steps or positions. For a tripeptide built from 20 amino acids at each of 3 positions, that is 20³ = 8,000 compounds. The exponent is always the number of steps, so adding a fourth position (20⁴ = 160,000) grows the library far faster than adding building blocks.
Why can mRNA display reach far larger libraries than phage display?
Phage display is capped by how many cells you can transform (around 10⁹ members), because each library member must be carried and amplified in a living cell. mRNA display (RaPID) does the linking chemically — puromycin ties a peptide to its own mRNA in vitro — so diversity is limited only by concentration and Avogadro's number, allowing libraries above 10¹⁴ and hits with drug-like (1–50 nM) affinity.
Why are cyclic peptides good for protein-protein-interaction targets but tricky as oral drugs?
Their macrocyclic shape presents a larger, more rigid surface than a typical small molecule, so they can engage the flat interfaces of PPI targets with high affinity. The same size and polarity, however, usually make them poorly permeable and not orally available, so cyclic-peptide leads often need delivery engineering or a non-oral route despite their attractive potency.
Can AI help me with the macrocycle and display-screening material?
Yes. Sia can drill the nˣ library-size calculation, compare phage versus mRNA display on diversity and affinity, explain how deconvolution or DNA barcoding identifies a hit, and place cyclic peptides in the beyond-rule-of-5 landscape. It explains the method and checks your reasoning; it does not do graded work for you, and University of Sydney academic-integrity rules apply.
Exam move
Make the nˣ relationship second nature — practise reading 'n building blocks over x steps' straight into an exponent, and notice how the library explodes with each extra step. Build a one-line comparison of the display platforms (phage ~10⁹, mRNA/RaPID >10¹⁴) with their typical hit affinities, and be able to say why the more diverse method reaches better potency. Keep the deconvolution/encoding idea attached to split-pool: you can make nˣ compounds cheaply, but you must be able to identify the winner. When the platform numbers or linker chemistries blur, ask Sia to quiz you and to set fresh nˣ problems.
Working through Macrocycles & Massively-Parallel Screening in MCHM3001? Sia is AskSia’s AI Chemistry tutor — ask any MCHM3001 Macrocycles & Massively-Parallel Screening question and get a clear, step-by-step explanation grounded in how MCHM3001 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
