Computational genomics and single-molecule biophysics · BEGINNER LESSON
Sequence-to-function
Sequence-to-function refers to methods that link a biological sequence, such as DNA or RNA, to a measured functional outcome. In practice, it often means a computational model that uses sequence to predict experimental signals or molecular behavior.
What it is and what is measured
Sequence-to-function methods ask how a sequence relates to a measured result. In genomics, the measured result can be a genome track: a position-by-position experimental signal for gene expression, RNA processing, DNA accessibility, histone modification, transcription factor binding, or chromosome contact. In single-molecule studies, the measured result can be real-time behavior such as folding, moving, or switching between structural states. The central goal is to connect the identity of a sequence to an observable function instead of treating the sequence as isolated text.
[1] From sequence to function: Bridging single-molecule kinetics and molecular diversity [2] Advancing regulatory variant effect prediction with AlphaGenome [3] Context-aware sequence-to-function model of human gene regulation
How the connection is learned
One major approach uses deep learning. A model is trained on many examples where the input is DNA sequence and the output is an experimental track or other functional measurement. After training, the model can predict outputs for new sequences and compare a reference sequence with an altered sequence to estimate variant effects. Another approach uses multiplexed single-molecule experiments: the motion or reaction pattern of many individual molecules is recorded first, and the exact sequence of each molecule is read out afterward, so sequence can be tied to a kinetic behavior profile.
[1] From sequence to function: Bridging single-molecule kinetics and molecular diversity [2] Advancing regulatory variant effect prediction with AlphaGenome
Why this matters biologically
Small sequence changes can alter structure, mechanics, binding, and reaction pathways. That can change folding, equilibrium between states, or the step-by-step cycles of molecular machines. For gene regulation, sequence-to-function models can predict signals linked to expression, chromatin accessibility, histone marks, splicing, and 3D chromatin contact. For molecules studied one by one, the same idea helps reveal rare intermediates and alternative pathways that bulk, population-averaged assays can hide.
[1] From sequence to function: Bridging single-molecule kinetics and molecular diversity [2] Advancing regulatory variant effect prediction with AlphaGenome [3] Context-aware sequence-to-function model of human gene regulation
Limitations
These methods still face important limits. Some models must trade off long-range context against fine base-level detail: short inputs can miss distant regulatory elements, while longer inputs may blur small features. Context also matters; a sequence may behave differently in different cell types or conditions, so sequence alone may not generalize well. On the experimental side, single-molecule multiplexing is limited by field of view, photobleaching, incomplete cluster formation, and difficulty capturing rare or unsynchronized reactions.
[1] From sequence to function: Bridging single-molecule kinetics and molecular diversity [2] Advancing regulatory variant effect prediction with AlphaGenome [3] Context-aware sequence-to-function model of human gene regulation
SEE THE IDEA
How sequence-to-function links information to outcome
DNA, RNA, or protein order provides the starting input.
Experimental signals or single-molecule dynamics are recorded as the output.
Learning algorithms or multiplexed experiments connect input to output.
Predicted regulation, dynamics, or variant effect is inferred from the link.
Sequence is connected to a measured functional readout through a model or an experiment.
The core idea is to pair molecular identity with an observable result, then interpret how changes in sequence alter that result.
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.