Among 6 Cysteine-Reference Substitution Pairs in ClinVar Missense Variants With ≥100 Records: Cys→Trp Is the Most Pathogenic-Enriched (75.1% Pathogenic, Wilson 95% CI [70.4, 79.2]) and Cys→Ser Is the Least (57.9% [54.6, 61.1]) — A 1.30× Range Within a Uniformly High-Pathogenicity Cysteine Substitution Set

Abstract

We compute the per-substitution-target-amino-acid Pathogenic fraction for the 6 Cysteine-reference (Cys, C) substitution pairs with ≥100 ClinVar missense single-nucleotide variants in the dbNSFP v4 (Liu et al. 2020) annotation of 372,927 ClinVar Pathogenic+Benign records (Landrum et al. 2018) returned by MyVariant.info (Wu et al. 2021), with Wilson 95% confidence intervals (Wilson 1927). Stop-gain (aa.alt = X) is explicitly excluded. Cysteine has the smallest per-target-AA Pathogenic-fraction range across reference AAs we have analyzed: only 6 (C → other) pairs exceed the ≥100-record threshold, and all 6 have Pathogenic fraction ≥ 0.57 — Cys substitutions are uniformly Pathogenic-enriched relative to the corpus-level baseline of ~28% Pathogenic. Result: per-target-AA Pathogenic fractions span a 1.30× range from 57.9% (C → S) to 75.1% (C → W): C→W 75.1% Wilson CI [70.4, 79.2]; C→F 69.3% [65.7, 72.6]; C→R 68.1% [65.8, 70.4]; C→G 64.6% [60.3, 68.7]; C→Y 63.8% [61.6, 66.0]; C→S 57.9% [54.6, 61.1]. The narrow per-pair range and uniformly-high Pathogenic fractions reflect the near-universal disruption caused by losing a cysteine: in folded proteins, Cys residues participate in disulfide bonds (Sevier & Kaiser 2002), metal coordination (Zn²⁺, Fe²⁺), or thiol-redox active sites; substitution of any other amino acid removes these functional roles. The chemistry of the alt AA modulates the magnitude (W > F > R > G > Y > S), but the range is tighter than for less-functionally-constrained reference AAs (Arg 4.2× range, Gly 2.2× range from companion analyses). The C → S substitution at 57.9% Pathogenic (the most-Benign Cys-derived pair) reflects that Ser is the closest amino acid to Cys in chemistry (sulfhydryl → hydroxyl), preserving partial H-bonding capacity at the cost of disulfide formation. For variant-prioritization pipelines: any C → X substitution carries a > 50% Pathogenic prior (the lowest C → S still at 57.9%), substantially above the corpus-baseline 28%.

1. Background

Cysteine (Cys, C) is a sulfur-containing amino acid with a unique reactive side chain (-CH₂-SH). In folded human proteins, Cys residues participate in three primary functional roles:

• Disulfide bond formation (-S-S-): the most common Cys role in extracellular and secreted proteins (Sevier & Kaiser 2002). Disulfides are stabilizing covalent crosslinks essential for tertiary fold.
• Metal coordination: Cys-rich motifs (CXXC, CCCC, zinc fingers) coordinate Zn²⁺, Fe²⁺, Cu²⁺.
• Active-site thiol chemistry: catalytic Cys residues in proteases (caspases, papain), kinases, and oxidoreductases.

All three roles require the specific Cys -SH side chain; substitution of any other amino acid removes the function. This paper measures the per-target-AA Pathogenic-fraction distribution for Cys-reference substitutions.

2. Method

Identical to companion per-AA-substitution-pair analysis: ClinVar missense (alt ≠ X) variants from MyVariant.info / dbNSFP v4. Restrict to ref = C; group by alt AA; require ≥100 total per pair. Wilson 95% CI on the per-pair Pathogenic fraction.

The "Mean rel pos" column reported in result.json is the per-pair mean of aa.pos / protein_length across all Pathogenic variants in the pair — a centrality summary of where in the protein these substitutions occur. We report it for completeness; for the Cys-reference set, all per-pair mean relative positions cluster around 0.43–0.48 (slight N-terminal skew), with no per-pair difference larger than 0.05. The mean-position values do not drive the per-pair P-fraction differences.

3. Results

3.1 Per-target-AA Pathogenic fraction (sorted descending)

3.2 The narrow per-pair range and uniformly-high Pathogenicity

Cys-reference Pathogenic fractions are uniformly above 57%, with the highest (C → W) reaching 75.1%. The 1.30× range (75.1 / 57.9) is narrower than for Arg-reference (4.2× range, 15.0% to 63.1% from companion) and Gly-reference (2.2× range, 28.9% to 63.7% from companion).

The mechanism: Cys is functionally constrained at most positions where it appears (disulfide bonds, metal coordination, active sites). Substitution of any other amino acid removes the disulfide / metal-coordination / catalytic role, with Pathogenic consequence. The chemistry of the alt AA modulates the magnitude but doesn't determine whether the substitution is consequential.

3.3 The chemistry-class ranking

Tier 1 — Most-Pathogenic Cys substitutions (P-fraction > 65%):

• C → W (75.1%): introduces large bulky aromatic side chain in place of small reactive thiol. Maximum volume increase among Cys-derived pairs.
• C → F (69.3%): introduces aromatic side chain. Less volume-disruptive than W but no thiol functional replacement.
• C → R (68.1%): introduces large basic side chain. Charge introduction at typically uncharged Cys positions.

Tier 2 — Mid-range Cys substitutions (P-fraction 60–65%):

• C → G (64.6%): introduces conformational flexibility (Gly is the smallest AA). Disrupts structural roles by removing rigidity.
• C → Y (63.8%): introduces aromatic ring with hydroxyl (intermediate between W and F in volume; preserves hydroxyl H-bonding).

Tier 3 — Most-Benign Cys substitution (P-fraction < 60%):

• C → S (57.9%): replaces sulfhydryl (-SH) with hydroxyl (-OH). Preserves H-bonding capacity, polar character, and approximate side-chain volume. The chemistry is the closest replacement for Cys among the 19 alternatives. Yet still 57.9% Pathogenic — Cys substitution is rarely benign even with the chemically-closest alt AA.

3.4 The C → S "least-Pathogenic" caveat

C → S at 57.9% Pathogenic is the most-Benign Cys-derived substitution but is still substantially above the corpus-baseline 28% Pathogenic fraction. Cys-reference variants are universally enriched in Pathogenic — the question is not "is this substitution likely Pathogenic" but "how strongly Pathogenic is it."

The -SH → -OH substitution preserves polarity and H-bonding but removes:

• Disulfide bond formation capability.
• Metal coordination (Cys-S binds Zn²⁺ ~10⁵× more strongly than Ser-O).
• Thiol-based catalysis.

For the ~58% of C → S variants that are Pathogenic, the position likely required one of these three functions. For the ~42% that are Benign, the Cys was likely a non-functional surface residue.

3.5 Mean relative positions are similar across pairs

All 6 Cys-derived pairs have mean relative position 0.43–0.48 (slight N-terminal skew). There is no per-pair position bias for Cys-reference Pathogenic variants. The N-terminal skew (~5% below the uniform 0.50) is consistent with the disulfide-bond-rich extracellular and secreted protein subset, which often has signal peptides or N-terminal domains containing the first disulfide bridges.

4. Confound analysis

4.1 Stop-gain explicitly excluded

We filter alt = X. Reported numbers are missense-only.

4.2 ClinVar curatorial bias and gene-level overrepresentation

Cys-rich functional motifs are well-studied in disease genes: zinc fingers (KLF, GLI, ZNF families), EGF-like repeats (NOTCH1, fibrillin/FBN1), kringle domains, immunoglobulin-fold domains, and laminin-like domains. Many ClinVar Cys Pathogenic variants come from these gene families. The uniformly-high Cys-reference Pathogenic fractions therefore partly reflect the functional-motif curation focus rather than a generic Cys-pathogenicity rule across all genes. A complementary analysis stratified by gene-family (e.g., disulfide-bond-rich extracellular vs cytoplasmic) would refine the per-Cys-position interpretation.

4.3 Codon-mutability not normalized

Cys has 2 codons (TGT, TGC). The per-target-AA mutational rates differ across the 6 alt AAs reported. C → R is achieved through TGN → CGN single-nucleotide transitions which are common; C → W requires TGN → TGG. We report the raw P-fraction observed in ClinVar.

4.4 Per-isoform first-element AA

We use the first finite element of dbnsfp.aa.ref and dbnsfp.aa.alt. ~5% per-isoform mismatch.

4.5 Wilson CI assumes binomial sampling

Per-pair counts are binomial. Wilson 95% CI is appropriate (Brown et al. 2001).

4.6 ACMG-PP3/BP4 partial circularity

ClinVar Pathogenic / Benign labels are partly predictor-derived (PolyPhen / SIFT scores used as PP3 evidence). Some per-pair fractions reflect predictor-curator co-variance.

4.7 N-threshold sensitivity

We use ≥100 total per pair. Cys-derived substitutions with < 100 records (C → A, C → V, C → L, C → I, etc.) are not analyzed. These substitutions require 2-step codon transitions and are infrequent.

5. Implications

1. Cys-reference Pathogenic fractions are uniformly above 57% across the 6 analyzed pairs, much higher than the corpus-baseline ~28%.
2. C → W is the most Pathogenic Cys substitution at 75.1% (Wilson CI [70.4, 79.2]).
3. C → S is the least Pathogenic Cys substitution at 57.9% [54.6, 61.1] — but still 2× the corpus baseline.
4. The 1.30× per-target-AA range within Cys-reference is narrower than Arg (4.2×) and Gly (2.2×) — Cys-reference is uniformly functionally-constrained.
5. For variant-prioritization pipelines: any C → X substitution should carry a > 50% Pathogenic prior; the per-target-AA chemistry modulates the magnitude (W > F > R > G > Y > S).

6. Limitations

1. Stop-gain excluded (§4.1).
2. Cys-rich functional-motif curation focus (§4.2) — uniform high Pathogenicity partly reflects research-bias toward disulfide-rich genes.
3. No codon-mutability normalization (§4.3).
4. Per-isoform first-element AA (§4.4).
5. N-threshold ≥ 100 (§4.7) excludes 2-step-codon-distance pairs.
6. ACMG-PP3 partial circularity (§4.6).

7. Reproducibility

• Script: analyze.js (Node.js, ~60 LOC, zero deps).
• Inputs: ClinVar P + B JSON cache from MyVariant.info.
• Outputs: result.json with per-target-AA counts, P-fractions, Wilson 95% CIs, mean relative positions.
• Verification mode: 6 machine-checkable assertions: (a) all P-fractions in [0, 1]; (b) Wilson CIs contain the point estimate; (c) all 6 reported pairs have N ≥ 100; (d) C→W P-fraction > 0.7; (e) C→S P-fraction > 0.5; (f) sample sizes match input.

node analyze.js
node analyze.js --verify

8. References

1. Landrum, M. J., et al. (2018). ClinVar. Nucleic Acids Res. 46, D1062–D1067.
2. Liu, X., Li, C., Mou, C., Dong, Y., & Tu, Y. (2020). dbNSFP v4. Genome Med. 12, 103.
3. Wu, C., et al. (2021). MyVariant.info. Bioinformatics 37, 4029–4031.
4. Wilson, E. B. (1927). Probable inference, the law of succession, and statistical inference. J. Am. Stat. Assoc. 22, 209–212.
5. Brown, L. D., Cai, T. T., & DasGupta, A. (2001). Interval estimation for a binomial proportion. Stat. Sci. 16, 101–133.
6. Sevier, C. S., & Kaiser, C. A. (2002). Formation and transfer of disulphide bonds in living cells. Nat. Rev. Mol. Cell Biol. 3, 836–847.
7. Richards, S., et al. (2015). ACMG/AMP variant interpretation guidelines. Genet. Med. 17, 405–424.
8. Cooper, D. N., & Krawczak, M. (1990). The mutational spectrum of single base-pair substitutions causing human genetic disease. Hum. Genet. 85, 55–74.
9. Berg, J. M., & Shi, Y. (1996). The galvanization of biology: a growing appreciation for the roles of zinc. Science 271, 1081–1085. (Cys-Zn coordination reference.)
10. Grantham, R. (1974). Amino acid difference formula to help explain protein evolution. Science 185, 862–864.