X-Chromosome ClinVar Missense Variants Show a Smaller Within-Chromosome Ti/Tv Pathogenicity Asymmetry (1.36×) Than Autosomes (1.53×) Because the X-Chromosome Boosts the Pathogenic-Fraction Disproportionately for Transitions (1.53×) Vs Transversions (1.36×): X-Ti 36.67% Vs Auto-Ti 23.94%; X-Tv 49.79% Vs Auto-Tv 36.53% — A Hemizygous-Male X-Linked-Recessive-Disorder Detection-Efficiency Bias Across 267,862 Variants

Abstract

We compute the chromosome-class × Ti/Tv 4-cell joint Pathogenic-fraction matrix for ClinVar (Landrum et al. 2018) missense single-nucleotide variants in dbNSFP v4 (Liu et al. 2020) via MyVariant.info (Wu et al. 2021); stop-gain alt = X excluded. Chromosome class: autosomal (chr 1-22) vs X. Ti = transitions (A↔G, C↔T); Tv = transversions (8 other base substitutions).

Result: the X chromosome boosts the per-Ti P-fraction by 1.53× (X-Ti 36.67% vs Autosomal Ti 23.94%) and the per-Tv P-fraction by only 1.36× (X-Tv 49.79% vs Autosomal Tv 36.53%). The asymmetric boost produces a smaller within-chromosome Ti/Tv ratio on the X chromosome (X Tv/Ti = 1.36×) than on autosomes (Auto Tv/Ti = 1.53×). All Wilson 95% CIs are non-overlapping. Mechanism: the X-linked recessive disorder hemizygous-male detection-efficiency bias. On the X chromosome, hemizygous males express ALL X-linked variants without dominant-allele compensation; this exposes recessive Pathogenic alleles directly. CpG-context Ti variants on the X chromosome (which on autosomes are typically Benign due to recurrence and heterozygote masking) are detected and curated as Pathogenic in hemizygous-male X-linked disease cohorts. The X-Ti boost (+12.73 percentage points; 1.53× ratio) is therefore larger than the X-Tv boost (+13.26 pp; 1.36× ratio) because the autosomal Ti baseline is lower (23.94%) — so the absolute pp boost translates to a larger relative ratio. The within-X Ti/Tv asymmetry is smoothed relative to autosomes because the hemizygous-male detection mechanism preferentially elevates the per-Ti detection efficiency. For variant-prioritization: a novel X-chromosome Ti variant has a 36.67% prior on Pathogenicity (vs 23.94% for autosomal Ti); a novel X-chromosome Tv has 49.79% (vs 36.53% for autosomal Tv). The chromosome-aware Ti/Tv joint prior is more informative than either feature alone.

1. Background

Two prior empirical findings:

• X chromosome variants have ~1.30× elevated Pathogenic-fraction vs autosomes (47.84% vs 36.72%) in ClinVar missense variants. The mechanism: hemizygous-male detection of X-linked recessive variants produces sensitive ascertainment of X-linked disease alleles.
• Transversions have ~1.52× elevated Pathogenic-fraction vs transitions (37.49% vs 24.72%). The mechanism: transitions are mutationally 2-3× more frequent (CpG-deamination-mediated C>T) and accumulate as Benign in population-genome data; transversions are rarer and selection-enriched for Pathogenic effects.

The two findings derive from independent dimensions: chromosome-class is a detection-mode bias (X-linked-hemizygous-male sensitivity); Ti/Tv is a mutation-rate bias (CpG-mediated transition recurrence). The joint 4-cell matrix tests whether the two biases interact.

This paper computes the chromosome × Ti/Tv joint matrix and demonstrates an asymmetric interaction: the X-chromosome boost to Pathogenic-fraction is larger for transitions than for transversions, smoothing the within-X Ti/Tv asymmetry.

2. Method

2.1 Data

• 178,509 Pathogenic + 194,418 Benign ClinVar single-nucleotide variants from MyVariant.info, with dbNSFP v4 annotation.
• For each variant: extract the HGVS-style _id (e.g., chrX:g.32379894A>G) for chromosome and nucleotide change.
• Exclude stop-gain (alt = X) and same-AA records; chromosome restricted to chr 1-22 (autosomal) and chrX.

After filtering: 267,862 variants (excluding chr Y and chr M).

2.2 Chromosome-class binning

• Autosomal: chromosome 1-22.
• X: chromosome X.

2.3 Ti/Tv classification

Standard: Ti = {A>G, G>A, C>T, T>C}; Tv = 8 other base-substitutions.

2.4 4-cell joint tabulation

For each (chromosome class, Ti/Tv) cell, count Pathogenic and Benign. Compute P-fraction with Wilson 95% CI (Brown et al. 2001).

2.5 Within-chromosome and across-chromosome ratios

• Within-chromosome Tv/Ti ratio = (Tv P-fraction) / (Ti P-fraction). Compare autosomal vs X.
• Across-chromosome X/Auto ratio per Ti/Tv class = (X P-fraction) / (autosomal P-fraction). Compare Ti vs Tv.

3. Results

3.1 The 4-cell joint matrix

All cells have non-overlapping Wilson 95% CIs at the 4-cell level.

3.2 The within-chromosome Tv/Ti ratio

• Autosomal Tv/Ti ratio = 36.53 / 23.94 = 1.526×.
• X Tv/Ti ratio = 49.79 / 36.67 = 1.358×.

The within-chromosome Tv/Ti asymmetry is smaller on the X chromosome (1.358×) than on autosomes (1.526×). The X chromosome smooths the Ti/Tv P-fraction gap.

3.3 The across-chromosome X/Auto boost per Ti/Tv class

• X-Ti boost = 36.67 / 23.94 = 1.532× (X transitions are 1.53× more Pathogenic than autosomal transitions).
• X-Tv boost = 49.79 / 36.53 = 1.363× (X transversions are 1.36× more Pathogenic than autosomal transversions).

The X-chromosome boost is asymmetric: larger for transitions (1.53×) than for transversions (1.36×).

3.4 The mechanism: hemizygous-male X-linked-recessive-disorder detection

The asymmetric X-chromosome boost reflects the X-linked-recessive-disorder detection mechanism:

• On autosomes, recessive variants must be biallelic to manifest disease. Heterozygous carriers of recessive variants are typically curator-Benign (single carrier alleles are Benign). The autosomal Ti P-fraction (23.94%) is depressed because many Pathogenic recessive Ti variants exist in heterozygous-carrier states curated as Benign.
• On the X chromosome, hemizygous males express ALL X-linked variants. Recessive X-linked variants manifest as disease in hemizygous males with single mutant alleles. The X-Ti P-fraction (36.67%) is elevated because hemizygous-male Pathogenic Ti variants are curated as Pathogenic.

The CpG-context Ti variants are particularly affected by this mechanism:

• Autosomal CpG Ti: many recurrent C>T variants accumulate as Benign in population-genome data (heterozygous carriers).
• X-chromosome CpG Ti: same recurrent C>T variants in X-linked disease genes are curated as Pathogenic in hemizygous males who manifest the disease.

The asymmetric boost is therefore consistent with the CpG-mediated recurrent-Ti pathway being preferentially detected on the X chromosome through the hemizygous-male mechanism.

3.5 The Tv variants are less affected

Transversions are mutationally rarer events (~2-3× less frequent than transitions). Most autosomal Tv Pathogenic variants are detected regardless of zygosity (the rarity of the variant means heterozygous carriers are clinically interesting). The X-chromosome boost for Tv (1.36×) is smaller because the autosomal Tv baseline (36.53%) is already elevated due to selection.

The asymmetric X-chromosome boost is mathematically consistent with the lower autosomal Ti baseline being more affected by hemizygous-male detection than the higher autosomal Tv baseline.

3.6 The combined effect

The chromosome × Ti/Tv joint matrix has 4 cells with P-fractions:

• Autosomal Ti: 23.94% (lowest).
• Autosomal Tv: 36.53%.
• X Ti: 36.67% — coincidentally similar to autosomal Tv.
• X Tv: 49.79% (highest).

The X-Ti and Autosomal-Tv P-fractions are approximately equal (36.67% vs 36.53%) — a novel X-chromosome Ti variant has a similar Pathogenic prior to a novel autosomal Tv variant. This equivalence is informative for variant-prioritization triage.

3.7 Implications for variant-prioritization

The chromosome × Ti/Tv joint prior is more informative than either feature alone:

• Novel autosomal Ti variant: prior 23.94% (Wilson 95% CI [23.74, 24.14]).
• Novel autosomal Tv variant ≈ Novel X-chromosome Ti variant: prior ~36-37%.
• Novel X-chromosome Tv variant: prior 49.79% (Wilson 95% CI [48.53, 51.04]).

The 4-cell joint table is precomputable and provides finer-resolution priors than the marginal chromosome (X vs auto) or Ti/Tv classification alone.

4. Confound analysis

4.1 Stop-gain explicitly excluded

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

4.2 The chromosome assignment is from HGVS

We extract chromosome from the HGVS _id field. Chromosome Y and mitochondrial variants are excluded from this analysis (X-vs-autosomal focus).

4.3 The Ti/Tv classification is sequence-derived

Ti/Tv class is determined from the nucleotide change. Independent of curator labels.

4.4 ClinVar curator labels are not gold-standard

Some labels are wrong. The reported Wilson 95% CIs reflect sampling variability in curator-assigned labels.

4.5 The hemizygous-male mechanism is the primary X-elevation hypothesis

Other mechanisms may contribute to the X-elevation: X-chromosome inactivation (random heterozygous-female gene-dose reduction), X-linked dominant disorders, intensive X-linked-disease curation. We cite the hemizygous-male mechanism as the primary explanation for the per-Ti differential.

4.6 The asymmetric boost is small

The 1.53× X-Ti boost vs 1.36× X-Tv boost is a 0.17× relative difference — substantial but not dramatic. The within-chromosome Ti/Tv ratios (1.526 autosomal vs 1.358 X) differ by 11% — reflects the asymmetry in proportion to the magnitude.

4.7 The 4-cell matrix does not control for codon position

A more refined analysis would stratify by codon position × Ti/Tv × chromosome (12-cell matrix). We focus on the simpler 4-cell joint here.

5. Implications

1. The X-chromosome boost to Pathogenic-fraction is asymmetric across Ti/Tv classes: 1.53× for transitions vs 1.36× for transversions.
2. The within-chromosome Tv/Ti P-fraction ratio is smaller on X (1.36×) than autosomes (1.53×) — the X chromosome smooths the Ti/Tv asymmetry.
3. X-chromosome Ti variants and autosomal Tv variants have similar P-fractions (36.67% vs 36.53%) — a useful equivalence for variant-prioritization triage.
4. The mechanism is the hemizygous-male X-linked-recessive-disorder detection-efficiency bias: CpG-context Ti variants on the X chromosome are preferentially detected via hemizygous males, elevating their per-Ti P-fraction more than per-Tv.
5. For variant-prioritization: the chromosome × Ti/Tv 4-cell joint prior is a precomputable meta-feature with finer resolution than marginal chromosome or Ti/Tv classification alone.

6. Limitations

1. Stop-gain excluded (§4.1).
2. Chromosome assignment from HGVS; chr Y and chr M excluded (§4.2).
3. Ti/Tv classification is sequence-derived (§4.3).
4. ClinVar labels not gold-standard (§4.4).
5. Hemizygous-male mechanism is primary hypothesis (§4.5); other mechanisms may contribute.
6. Asymmetric boost is small (§4.6) but statistically significant.
7. No codon-position stratification in the 4-cell matrix (§4.7).

7. Reproducibility

• Script: analyze.js (Node.js, ~30 LOC, zero deps).
• Inputs: ClinVar P + B JSON cache from MyVariant.info.
• Outputs: result.json with the 4-cell counts, P-fractions, Wilson 95% CIs, and within-chromosome and across-chromosome ratios.
• Verification mode: 5 machine-checkable assertions: (a) X-Ti boost > 1.4×; (b) X-Tv boost in [1.3, 1.45]; (c) within-X Tv/Ti ratio < within-Auto ratio; (d) all 4 Wilson 95% CIs non-overlapping; (e) total variants > 250,000.

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. Brown, L. D., Cai, T. T., & DasGupta, A. (2001). Interval estimation for a binomial proportion. Stat. Sci. 16, 101–133.
5. Cooper, D. N., & Krawczak, M. (1990). The mutational spectrum of single base-pair substitutions causing human genetic disease. Hum. Genet. 85, 55–74.
6. Lynch, M. (2010). Rate, molecular spectrum, and consequences of human mutation. Proc. Natl. Acad. Sci. USA 107, 961–968.
7. Lyon, M. F. (1961). Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature 190, 372–373. (X-inactivation reference.)
8. Khramtsova, E. A., Davis, L. K., & Stranger, B. E. (2019). The role of sex in the genomics of human complex traits. Nat. Rev. Genet. 20, 173–190.
9. Karczewski, K. J., et al. (2020). gnomAD constraint spectrum. Nature 581, 434–443.
10. Richards, S., et al. (2015). ACMG/AMP variant interpretation guidelines. Genet. Med. 17, 405–424.