LIVIA — FlyPredictome Analysis

🔖 On a FlyPredictome details page? Jump straight to this analysis with the LIVIA bookmarklet — one click, no copy-paste. Get it →

FlyPredictome Interaction Lookup

Precomputed PPI predictions for Drosophila from FlyPredictome. Search by gene name or paste a prediction page URL.

Examples:

Interaction Analysis Summary

↓ Download CSV

Ranks are FlyPredictome's own order of the five models, which is ipTM-based, not iLIS-based. Click a row to switch ranks; this updates all visualizations and the script below.

Rank	Symbol 1	Symbol 2	iLIS	iLIA	iLISA	ipTM	pDockQ	LIpDockQ	LIR (A/B)	cLIR (A/B)	Len (A/B)

iLIS: ≥0.551 (1% FPR) ≥0.339 (5% FPR) ≥0.223 (10% FPR) <0.223 (below threshold)

iLIS (average): ≥0.268 (1% FPR) ≥0.120 (5% FPR) ≥0.072 (10% FPR) <0.072 (below threshold)

Thresholds based on large-scale Y2H reference sets in yeast, fly, and human predicted using ColabFold. See Kim et al., 2026, for iLIS benchmark details. Rank 1 (determined by ipTM) may not have the highest iLIS.

Reported Interactions (MIST, BioGRID)

Predicted Aligned Error (PAE)

PAE matrix showing prediction confidence between all residue pairs. Blue = low error (high confidence).

Sequence Viewer

Light background = LIR (PAE ≤ 12 Å). Dark background = cLIR (PAE ≤ 12 Å & Cβ ≤ 8 Å). Each chain uses its own color.

Linear Contact Map

Lines connect cLIR residue pairs within Cβ distance in the 3D structure. Per-rank structures fetched automatically.

domains: UniProt/Pfam TED pLDDT chain IDs swap chains (top↔bottom) x-ticks

Circular Contact Map

labels chain IDs length ticks pLDDT domains: UniProt/Pfam TED label direction

Arcs represent chains (proportional to length). Lines connect cLIR residue pairs in physical contact.

Selected Rank

3D Structure Viewer

Interactive 3D preview powered by Mol* (Sehnal et al. 2021). LIR regions shown as cartoons; cLIR residues highlighted. Drag to rotate, scroll to zoom.

Visualization Script

Local Interaction Residue (LIR) Display

Fill gaps ≤ residues | Min segment ≥ residues

Gap filling bridges short breaks for continuous cartoon. Min segment removes isolated LIR fragments shorter than the threshold.

Show complete structure (display all residues instead of LIR only) Gray non-LIR (color non-interacting residues gray) Color Scheme

Chain A (Protein 1)

LIR — confident domain (PAE ≤ 12)

#80CBC4

cLIR — confident interface (PAE ≤ 12 & Cβ ≤ 8 Å)

#00897B

Chain B (Protein 2)

LIR — confident domain (PAE ≤ 12)

#FFAB91

cLIR — confident interface (PAE ≤ 12 & Cβ ≤ 8 Å)

#E64A19

Presets — gradient (LIR light, cLIR dark)

Teal / Coral

Blue / Orange

Purple / Gold

Slate / Rose

Indigo / Tangerine

Green / Red

Presets — high contrast (all 4 colors distinct)

Cyan / Navy / Peach / Crimson

Sky / Indigo / Salmon / Red

Lime / Forest / Gold / Maroon

Mint / DarkTeal / Apricot / Burnt

Orchid / Purple / Yellow / Fire

Blue / Navy / Sage / Ember

Presets — solid

Teal / Coral

Blue / Orange

Purple / Gold

Slate / Rose

Steel / Wheat

Cornflower / Tomato

SkyBlue / Gold

LightGreen / Plum

Salmon / Teal

Presets — coloring modes

pLDDT

bychain

bypolymer

↓ Full structure .pdb ↓ .cxc · for full ↓ Partial structure .pdb ↓ .cxc · for partial

Full structure .pdb = full length (surfaces include the floppy N/C-terminal tails). Partial structure .pdb = both chains trimmed to exactly the LIR segments shown (Fill gaps / Min segment honored) — clean surfaces for post-processing; residue numbers preserved so the script still colors correctly. Download one structure file.

The .pdb link opens the file from FlyPredictome. Save it in the same folder as the script file, then open the script in ChimeraX or PyMOL.

Per-Residue pLDDT

Per-residue pLDDT for each chain across all ranks. Selected rank as solid line; other ranks as avg ± SD shading.

How to Use (Full guide)

Quick Start

Go to FlyPredictome and search for a protein pair
Copy the FlyPredictome result page URL
Paste the URL above and click Generate
Download the script (.cxc or .pml) and structure file, place both in the same folder
Open the script in ChimeraX (.cxc) or PyMOL (.pml)

AFM-LIS Metrics (Kim et al. 2024; Kim et al. 2026)

iLIS	integrated LIS — √(LIS × cLIS)
iLIA	integrated LIA — √(LIA × cLIA), geometric mean of interface area counts
iLISA	integrated LISA — iLIS × iLIA, overall binding strength
LIS	Local Interaction Score — normalized PAE confidence (0–1)
cLIS	contact-filtered LIS — restricted to direct contacts
LIR	Local Interaction Residues (PAE ≤ 12 Å)
cLIR	contact-filtered LIR (PAE ≤ 12 Å & Cβ ≤ 8 Å)

Confidence Metric (from ColabFold)

ipTM

interface predicted TM-score — global interface confidence from the prediction model

pDockQ family

pDockQ	predicted DockQ from the mean pLDDT of the interface and the number of contacting residue pairs (Cβ ≤ 8 Å), with no PAE (Bryant et al., 2022)
LIpDockQ	the pDockQ formula computed only on cLIR pairs (PAE ≤ 12 Å and Cβ ≤ 8 Å), so contacts the PAE does not support no longer count (Kim & Perrimon, 2026)

pDockQ2 and LIpDockQ2 need the numeric PAE, which this source does not provide. Because their constants were fitted to DockQ on the original interfaces, LIpDockQ and LIpDockQ2 rank predictions; they are not DockQ estimates.

PAE (Predicted Aligned Error) measures how confidently the model predicts the relative position of two residues — lower values indicate higher confidence.