Aggregating Rare Variants to Uncover Missing Heritability

You are reading version 1 of this direction, frozen 2026-10-04. The latest version is version 2.

Version 1 of 2118 papers150 conceptsfrozen 2026-10-04

In the search for the genetic roots of complex diseases, researchers often face a "signal-to-noise" problem. Standard Genome-Wide Association Studies (GWAS) are excellent at finding common variants—mutations shared by a large percentage of the population. However, much of the "missing heritability" of diseases like schizophrenia, ALS, and diabetes is thought to reside in rare variants: mutations that appear in fewer than 1% (or even 0.1%) of the population.

Individually, these rare variants are statistically "quiet." If only three people in a study of 10,000 carry a specific mutation, no statistical test can reliably link that mutation to a disease. To solve this, researchers have shifted from testing individual markers to aggregating them into functional units.

The Intuition of Collapsing

The core idea, often called rare variant collapsing or burden testing, is to treat a group of rare mutations as a single statistical unit. As Pingzhao Hu and Andrew D. Paterson describe it, this approach converts individual "whispers" from infrequent variants into a collective "shout" that a statistical model can actually hear.

Instead of asking, "Is SNP A associated with the disease?", researchers ask, "Does this gene contain a higher-than-expected number of damaging mutations in patients compared to healthy controls?" By grouping variants within a gene, researchers like Nicholas P. Tatonetti and Atul Butte have shown that you can significantly reduce the "multiple testing burden"—the statistical penalty for checking millions of hypotheses—thereby increasing the power to find real associations in smaller cohorts.

Burden Tests vs. Kernel Tests

There are two primary ways to aggregate these signals, each suited for different biological realities:

  1. Burden Tests: These methods, such as those evaluated by Wenan Chen and Nianjun Liu, sum up the number of rare variants in a genomic region. This works best when all mutations in that region affect the trait in the same direction (e.g., all of them increase the risk of a heart rhythm disorder, as explored by Elijah R. Behr).
  2. Sequence Kernel Association Tests (SKAT): In many cases, some mutations in a gene might be harmful while others are protective. A simple sum would cancel these effects out. SKAT, used by Yuanlin Ma and Hongxu Pan, evaluates the variance of the effects rather than the mean. This allows it to detect associations even when the direction of the mutations' impact is mixed.

Scaling to Biobanks

As genetic datasets have grown to include hundreds of thousands of individuals, the computational challenge has shifted. Tools like SAIGE-GENE+, developed by Wei Zhou and Wenjian Bi, use binary indicators to handle "ultra-rare" variants. If a variant is so rare that it only appears once or twice in a massive biobank, SAIGE-GENE+ collapses these into a simple "present/absent" flag for the gene, which stabilizes the math and prevents false positives.

This scalability is what allowed Parsa Akbari and Jonas B. Nielsen to analyze over 600,000 exomes, identifying rare loss-of-function mutations in the INHBE gene that protect against Type 2 Diabetes.

Beyond the Gene

While the gene is the most common unit of aggregation, the principle is flexible. Researchers have successfully applied burden testing to:

  • Biological Pathways: Grouping variants across multiple genes that work together in a single process, like vitamin C metabolism Ryan Swan.
  • Structural Variants: Aggregating rare deletions or duplications (CNVs) within specific cytogenetic bands to explain brain structure differences in schizophrenia A Pilot Study on Collective Effects of 22q13.31 Deletions on Gray Matter Concentration in Schizophrenia.
  • Imaging Phenotypes: Linking the collective burden of mutations to physical "chaos" seen in tumor CT scans Dandan Zheng.

By moving from a "one-variant-at-a-time" mindset to a "burden-of-variants" framework, researchers are finally beginning to map the complex, low-frequency genetic landscape that defines human health and disease.

Papers in this version

Showing 60 of 118 papers.

History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 74 papers
  2. Version 1 · 2026-10-04 · 118 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_rare_variant_aggregation_bd9d6ce7,
  title = {rare variant aggregation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/rare-variant-aggregation/rdr_bd9d6ce7a23a480382e807e41b6ba6f6},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_aggregating_rare_variants_to_uncover_mis_bd9d6ce7_v1,
  title = {Aggregating Rare Variants to Uncover Missing Heritability},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/aggregating-rare-variants-to-uncover-missing-heritability/rdr_bd9d6ce7a23a480382e807e41b6ba6f6/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}