# Hotspots and numbering

Use author residue numbers when defining hotspots for each model.

## Author numbering

A residue is named the way the deposited structure names it: chain, then the author's residue number, as in `A150`. That number is not its position in the sequence. Chains routinely start above 1, skip numbers at gaps, and carry insertion codes. Phi keeps author numbering everywhere a residue is shown or typed, and converts only at the edge where a model wants something else.

## The chain summary

Loading a target returns a chain summary before anything runs: chain ids, lengths, the numbering range, and the sequence. Read it before choosing hotspots. The example below is TFR1 chain A from PDB 9GH7, whose first residue is 122.

_Interactive: NumberingDemo. See https://phi.dynotx.com/docs/guides/hotspots-and-numbering._

## Avoid shifted hotspot positions

A hotspot list written as sequence positions, or a structure renumbered from 1, aims the run at the wrong residues without any error. On a chain that starts at 122, position 29 and residue 150 are the same atom; a model given `29` as an author number would look 121 residues away. Phi's hotspot fields show the chain's numbering range beside them for this reason.

## What each model takes

Most generators take hotspots as a list of author-numbered residues under `params.hotspots`. Some take them elsewhere or in another form. The table shows one set as each model wants it; the values are computed from the target's real numbering.

_Interactive: HotspotFormatTable. See https://phi.dynotx.com/docs/guides/hotspots-and-numbering._

Through the MCP tools, `phi_design_binder` takes the author-numbered list and does the conversion. Through the REST API and the CLI you write the model's own field.

---
Source: https://phi.dynotx.com/docs/guides/hotspots-and-numbering