Biotechnology

What Is Bioinformatics?

Bioinformatics uses computing, statistics, and software to store, analyze, and interpret biological data such as DNA sequences and protein structures. It turns raw genomic data into findings that can guide drug discovery.

Key points

  • Bioinformatics combines biology, computer science, and statistics to make sense of large biological datasets.
  • Core tasks include aligning DNA sequences, identifying genetic variants, measuring gene activity, and predicting protein structure.
  • BLAST, a sequence-search tool published in 1990 by researchers at the National Center for Biotechnology Information, remains one of the most widely used programs in biology.
  • AlphaFold, an AI system that predicts protein structures from their amino acid sequences, shared in the 2024 Nobel Prize in Chemistry.
  • In drug discovery, bioinformatics links genetic data to disease to find and prioritize drug targets.

A single human genome generates hundreds of gigabytes of raw sequencing data. A large study may sequence hundreds of thousands of people. None of that data means anything until it is processed, compared, and interpreted. Bioinformatics is the field that does this work, and it sits underneath almost every modern discovery in genomics and drug development.

The definition

Bioinformatics is the use of computing, mathematics, and statistics to collect, store, analyze, and interpret biological data. The data can be DNA and RNA sequences, protein structures, gene activity measurements, or clinical records. Bioinformatics overlaps with computational biology, which focuses more on building models of how biological systems work. In practice the two terms are often used interchangeably.

What bioinformaticians actually do

Task What it involves Why it matters
Sequence alignment Matching DNA or protein sequences against each other or a reference Identifies genes, variants, and evolutionary relationships
Variant calling Finding where a person's DNA differs from the reference genome The basis of genetic diagnosis and population genetics
Variant interpretation Predicting and classifying the effect of each variant Separates disease-causing variants from harmless ones
Gene expression analysis Measuring which genes are active in RNA data Reveals what changes in disease and in response to drugs
Genome-wide association studies Testing millions of variants for links to a disease or trait Finds genes and pathways involved in common diseases
Structural bioinformatics Predicting and analyzing the 3D shapes of proteins Guides the design of drugs that fit a target
Data integration Combining genetic, protein, and clinical data Connects molecular findings to patient outcomes

A typical genomics pipeline

When a genome is sequenced with next-generation sequencing, bioinformatics takes over from the moment reads come off the machine:

  1. Quality control: checking and filtering the raw reads.
  2. Alignment: placing each short read at its correct location on the reference genome.
  3. Variant calling: comparing the aligned reads with the reference to find differences.
  4. Annotation: labeling each variant with what is known about it, such as which gene it falls in, how common it is, and whether it changes a protein.
  5. Interpretation: deciding which variants are meaningful for the question being asked.

Landmark tools

BLAST

In 1990, researchers at the National Library of Medicine's National Center for Biotechnology Information (NCBI) published the Basic Local Alignment Search Tool (BLAST). BLAST rapidly searches a sequence against large databases to find similar sequences, an order of magnitude faster than comparable tools at the time, according to the original paper. It remains one of the most widely used programs in biology. NCBI also maintains many of the public databases that bioinformatics depends on, including GenBank, PubMed, and ClinVar.

AlphaFold

A protein's function depends on its three-dimensional shape, and determining that shape experimentally can take years. In 2021, researchers at DeepMind reported in Nature that their AI system AlphaFold could predict protein structures from amino acid sequences with high accuracy. Demis Hassabis and John Jumper shared the 2024 Nobel Prize in Chemistry for protein structure prediction, together with David Baker, who was recognized for computational protein design. Predicted structures for a very large number of proteins are freely available through the AlphaFold Protein Structure Database.

The GWAS Catalog

The NHGRI-EBI GWAS Catalog collects results from published genome-wide association studies, linking genetic variants to diseases and traits. It is one of the main resources used to check whether a potential drug target has human genetic support.

Bioinformatics in drug discovery

Bioinformatics is now involved at nearly every early stage of drug development:

  • Target identification. Searching genetic association data for genes linked to a disease. Drug mechanisms with human genetic support have been about 2.6 times more likely to succeed in clinical development, according to a 2024 analysis in Nature.
  • Target assessment. Checking where a target protein is found in the body, which predicts both effectiveness and side effects.
  • Structure-based design. Using experimental or predicted protein structures to design molecules that fit a target.
  • Biomarker discovery. Analyzing patient data to find markers that predict who will respond to a drug.
  • Drug repurposing. Searching data for approved drugs whose effects might help other diseases.

Limitations

  • Garbage in, garbage out. Analyses are only as good as the underlying data, which can contain errors, batch effects, and gaps.
  • Biased reference data. Many genetic databases over-represent people of European ancestry.
  • Predictions are not proof. A predicted structure, variant effect, or target still needs experimental confirmation.
  • Reproducibility. Results can depend on software versions and analysis choices, so good practice requires sharing code and parameters.

Sources

  1. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. Journal of Molecular Biology. 1990;215(3):403–410.Peer-reviewed
  2. Jumper J, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596:583–589.Peer-reviewed
  3. The Nobel Prize in Chemistry 2024: press release.Industry report
  4. National Center for Biotechnology Information (NCBI).Registry
  5. NHGRI-EBI GWAS Catalog.Registry
  6. Minikel EV, Painter JL, Dong CC, Nelson MR. Refining the impact of genetic evidence on clinical success. Nature. 2024;629:624–629.Peer-reviewed

This article is for education. It is not medical advice and not investment advice. Figures are drawn from the sources above and describe historical averages, not the prospects of any particular drug. See our editorial standards and disclaimer.