Genomics and statistics · BEGINNER LESSON
Summary statistics
Summary statistics are calculated numbers that describe a study sample instead of listing each person’s raw measurements. In genomics, they can include genotype counts, allele frequencies, effect size estimates, standard errors, and p-values computed from many people’s genotypes.
What they are and what is measured
Summary statistics are measurements calculated from data after the raw observations have been combined. In genomic studies, they are often called genomic summary statistics and can include genotype counts, allele frequencies, effect size estimates, standard errors, and p-values. These values are derived from many individuals’ genotypes collected from arrays or sequencing, so they summarize a study rather than a single person. In simple terms, they measure how often variants appear, how strongly a variant is linked to a trait, and how uncertain that estimate is. [0,1]
[1] Background Value of Genomic Summary Statistics [2] Harmonisation Codes:
How they are used
Because they are compact, summary statistics can be shared and reused more easily than individual-level data. In genomics, they help compare one person’s genome or exome with large groups of others, which can improve interpretation of variants. Allele frequencies can suggest whether a variant is likely to be harmless or suspicious in a rare disease context, and p-values can help show which genomic regions are robustly associated with disease. Summary statistics can also be combined across studies for broader analysis. [0,2]
[1] Background Value of Genomic Summary Statistics [3] Multi-Task Learning with Summary Statistics - PMC
How they are organized and shared
Public databases may store summary statistics in standardized files so they can be reused. The GWAS Catalog uses full aggregate association data, and it prefers files that are freely available, large enough to be useful, and harmonized. Harmonization means reporting positions against the same genome build and aligning alleles to the same strand so different studies can be compared more reliably. [1]
Limitations and cautions
Summary statistics do not replace all raw-data analysis. Some methods can work with them, but one excerpt notes that inference may be less accurate than using individual-level data, especially when sample sources differ or overlap is limited. In genomics, privacy concerns also depend on what can be learned about whether a person participated in a study, so access and use may have limits. [0,2]
[1] Background Value of Genomic Summary Statistics [3] Multi-Task Learning with Summary Statistics - PMC
SEE THE IDEA
From raw genotype data to summary statistics
Raw variant data collected from arrays or sequencing.
Counts, frequencies, effect sizes, standard errors, and p-values are computed.
A compact table that describes the study sample.
Researchers compare studies, interpret variants, or combine results.
Raw genomic data are processed into compact study-level numbers, which are then reused for comparison and analysis.
The key idea is compression: many individual measurements become a smaller set of study-level values.
TRY IT YOURSELF
Test your understanding
Choose an answer, then check the explanation.
PUT IT TOGETHER
Match each concept to its meaning
Drag a label onto its description, or select a label and then select a description. This also works with a keyboard.
4 pairs to match
Sources and further reading
This general lesson was generated from public source material. It explains a concept; it does not interpret a private paper or provide medical advice.