WheatOmics 2.0 · Analysis tool

wheatPTM

PlantPTM deep-learning predictions of nine post-translational modification (PTM) types for wheat proteins. Search an exact protein ID to explore predicted modification sites, their confidence class and prediction scores.

— PTM calls — scored sites — proteins screened 9 PTM types

Protein lookup

Search by protein ID

Enter an exact protein ID — both standard gene-model proteins and LC models are indexed. Results list predicted sites with confidence classes and scores; Non-PTM predictions are hidden by default.

Example: TraesCS1A02G000100.1 · dataset overview ↓

Searching wheatPTM…

Dataset snapshot

Overview

Scored sites by modification type

Stacked bars: PTM calls (coloured by type) and Non-PTM predictions (grey). Hover a bar for counts and the PTM-call rate.

Confidence distribution

Across all scored sites. Non-PTM is the prediction tools’ negative class.

Example proteins

A few wheat proteins with moderate site counts — click to load.

Search result

Show Download CSV

PTM sites along the protein sequence

Scores are prediction scores from the source tools: N-glycosylation uses e-values (smaller = stronger), other types use 0–1 scores (larger = stronger). Scores of different PTM types are not directly comparable.

Protein	Position	Residue	PTM type	Peptide	Score	Confidence

About this dataset

Coverage: 9 PTM types — phosphorylation, N-glycosylation, S-acylation, 2-hydroxyisobutyrylation, crotonylation, succinylation, malonylation, acetylation and ubiquitination.
Predictions & confidence: rows are model predictions; Confidence is the source tool’s confidence class. Non-PTM rows are the negative class and are hidden by default — use the Show selector to include them.
Scores: tool-specific. N-glycosylation scores are e-values (smaller = stronger); other types report 0–1 scores (larger = stronger). Do not compare scores across PTM types.
Protein IDs: exact match only. Standard gene-model proteins (e.g. TraesCS1A02G000100.1) and LC protein models (e.g. TraesCS1A02G000100LC.1) are indexed separately.
Data source: predictions generated with PlantPTM (Dou et al., Molecular Plant 2026) and imported from its output CSV files (all_pho.csv, all_NGly.csv, …). The CSV download follows the current Show selection.
Interpretation: every row is a computational prediction, not an experimental observation — use the confidence classes when prioritising sites.

References & data sources

Dou G., Wang K., Qin Z., Zhang M., Zhang W., Miao C., Li J., Chen Z. PlantPTM: A deep-learning framework integrating protein language models with multi-view features for predicting diverse post-translational modification sites in plants. Molecular Plant 19(10): 1979–1995 (2026). doi: 10.1016/j.molp.2026.08.002
PlantPTM web server: https://ai4bio.online/PlantPTM · source code: https://github.com/wky0422/PlantPTM