TALOS-N prediction¶

makeshift.talosn wraps the NIH TALOS-N binary (Shen & Bax, J. Biomol. NMR 2013), which predicts backbone φ/ψ torsion angles, per-residue S² order parameters, and secondary structure from assigned backbone chemical shifts using a trained neural network.

The binary is downloaded on demand¶

TALOS-N and its database aren't bundled — they're downloaded from NIH (under their Terms of Use, which the installer prints) into a data_dir you choose.

Keep one data_dir

Store the path in a variable and pass the same data_dir to the installer and to every TalosN. If you omit it, it defaults to inside the installed package — usually not what you want for a few-hundred-MB download.

from pathlib import Path
from makeshift import talosn

data_dir = Path.home() / "talosn_data"
talosn.install_talosn_data(data_dir=data_dir)     # one-time, ~ a few hundred MB

You can check installation status with talosn.is_talosn_data_installed(data_dir).

Running a prediction¶

tn = talosn.TalosN.from_bmrb(4527, data_dir=data_dir)
tn.run()                    # or run(auto_install=True) to fetch on first use

tn.order_parameters         # predS2.tab — per-residue S2
tn.torsion_angles           # pred.tab   — phi/psi per residue + confidence class
tn.secondary_structure      # predSS.tab — helix / sheet / coil

Build from an already-parsed entry instead with TalosN.from_entry(entry, data_dir=data_dir). predict_s2() runs the pipeline if needed and returns the S² table directly.

Outputs¶

Property	TALOS-N file	Contents
`torsion_angles`	`pred.tab`	φ/ψ per residue + a prediction confidence class
`order_parameters`	`predS2.tab`	per-residue S²
`secondary_structure`	`predSS.tab`	helix / sheet / coil probabilities

Terms of use¶

The TALOS-N software is distributed separately by NIH under its own Terms of Use (including no redistribution without permission from the authors). Those terms govern the downloaded binary, not this wrapper.

Full API¶

See the TALOS-N reference.