LIVIA — About

1. Prediction Analysis

Quick Start

Run a prediction using any supported platform (AlphaFold3, AlphaFold2, ColabFold, Boltz-1/2, Chai-1, or OpenFold3)
Open Prediction Analysis and upload your prediction files (.zip, .gz, .cif, .pdb, .json, .npz, .npy)
The tool auto-detects the prediction platform and shows available models/ranks
Optionally adjust PAE cutoff (default: 12 Å) and Cβ cutoff (default: 8 Å), then click Process
Explore PAE/LIS/cLIS maps, score matrix, and chain pair table
Click a row in the chain pair table to generate a visualization script
Download the .cxc script and structure file, place them in the same folder, then open the script in ChimeraX or PyMOL

Supported Platforms

AlphaFold3 — .cif + summary_confidences + full_data JSON (from AlphaFold Server or local)
AlphaFold2 — ranked_*.pdb + PAE JSON (PAE must be pre-converted from .pkl)
ColabFold — *_unrelaxed_rank_*.pdb + *_scores_rank_*.json
Boltz-1/2 — .pdb/.cif + confidence_*.json + pae_*.npz
Chai-1 — pred.rank_*.cif + scores.rank_*.json (PAE in JSON or .npy)
OpenFold3 — result_sample_*_model.pdb + confidences JSON
Tamarind Bio — output from any platform via Tamarind (auto-detected)
Generic — any .cif/.pdb with a PAE JSON file

Features

Auto-detection — automatically identifies the prediction platform from file names
Compressed files — accepts .zip, .gz files; decompresses in browser
PAE maps — blue-white-red (bwr) colormap, per model/rank
LIS maps — matplotlib Blues colormap, PAE averaged both directions
cLIS maps — matplotlib Greens colormap, contact-filtered
Score matrix — iLIS (Oranges, lower-left) / ipTM (Purples, upper-right)
Residue count matrix — LIR (Blues) / cLIR (Greens)
Sequence viewer — amino acid letters with LIR/cLIR highlighting
Color presets — gradient, solid, pLDDT coloring, color bychain, color bypolymer
NPZ/NPY parser — reads Boltz PAE (.npz) and Chai-1 PAE (.npy) in browser
Downloads — .cxc script, structure file, CSV with full metrics

2. FlyPredictome Analysis

Quick Start

Go to FlyPredictome, search for a protein pair, and copy the prediction page URL
Open FlyPredictome Analysis, paste the URL, and click Generate
Review the All Ranks table — click any row to switch between ranks 1–5
Explore metrics (iLIS, ipTM, LIR/cLIR counts), the sequence viewer, and the visualization script preview
Download the script (.cxc for ChimeraX or .pml for PyMOL) and structure file, place them in the same folder, then open the script in ChimeraX or PyMOL

Features

All Ranks table — shows iLIS, LIS, cLIS, ipTM, LIR/cLIR counts for all 5 ranked models; click any row to load that rank
Sequence viewer — displays actual amino acid letters for each chain with LIR (light color) and cLIR (dark color) highlighting
Visualization script — syntax-highlighted preview with color presets (Gradient, Solid, coloring modes); Swap A ↔ B button reverses chain colors
Linear Contact Map — lines connect cLIR residue pairs in physical contact, with color gradient from chain A to chain B
Circular Contact Map — circular chord diagram showing chain arcs (proportional to length) and lines connecting cLIR residue pairs in contact
Downloads — .cxc script, .pdb structure, and CSV of all ranks with full metrics

3. Ortholog Interactome

Quick Start

Open Ortholog Interactome and search for a human gene (e.g. CDK2, TP53)
Browse the interaction partner table — click any row to load the detail page
Or paste a famdb_details_ortho URL directly in the Paste URL tab
Review ranks, contact maps, 3D viewer, and visualization script
Download the script and structure file for ChimeraX or PyMOL

Features

Gene search — UniProt autocomplete with species selector (Human, Mouse, Zebrafish, Worm, Yeast)
Partner table — shows iLIS (best/avg), ipTM, protein lengths; filterable by iLIS threshold
Pre-built examples — CDK2–CCNA2, TP53–MDM2, MAPK14–MKNK1, ABL1–CRK, and more
MIST integration — shows reported interactions from the MIST database
Full analysis pipeline — same as FlyPredictome (ranks, contact maps, Mol* viewer, scripts)

4. AlphaFold DB Dimer Analysis

Quick Start

Open AlphaFold DB Dimer Analysis and enter a UniProt ID (e.g. P69905), a model entity ID (e.g. AF-0000000066214167), or a full AlphaFold DB URL
Click Search Dimers — a table of available dimer complexes appears with LIS, ipTM, and ipSAE scores
Click a dimer row in the table to load it — the tool fetches CIF and PAE data from AlphaFold DB automatically
Review PAE/LIS/cLIS maps, score matrix, sequence viewer, and metrics (iLIS, LIS, cLIS, ipTM, ipSAE)
Download the script (.cxc/.pml) and structure file, place them in the same folder, then open the script in ChimeraX or PyMOL

Features

Flexible input — accepts UniProt ID, AlphaFold model entity ID, or full AlphaFold DB URL
Dimer search table — lists all available dimer predictions for a protein with complex name, type, gene names, LIS, ipTM, and ipSAE; click any row to select
Auto-fetch from AlphaFold DB — CIF structure and PAE matrix are fetched directly from the database
PAE/LIS/cLIS maps — same colormaps as the Universal tool (bwr, Blues, Greens)
Score matrix — iLIS (Oranges) / cLIR count (Greens)
Sequence viewer — amino acid letters with LIR (light) and cLIR (dark) highlighting per chain
ipSAE metric — interaction prediction Score from Aligned Errors (Dunbrack, 2025), shown alongside iLIS, LIS, cLIS, and ipTM
Visualization script — color presets with Swap A ↔ B support
Downloads — .cxc script, .cif structure, CSV with full metrics and LIR/cLIR indices

5. AlphaFold DB Monomer Subdomain Analysis

Quick Start

Open AlphaFold DB Monomer Subdomain Analysis and enter a UniProt ID (e.g. P31749) — autocomplete suggestions appear as you type
Click Fetch — the tool retrieves the PDB structure and PAE matrix from AlphaFold DB
Domains are auto-detected using pLDDT-based segmentation; adjust settings if needed (min length, merge LIS threshold, pLDDT cutoff)
Click Detect Domains to re-detect with new settings, or manually edit domain boundaries
Review PAE/LIS/cLIS maps, the iLIS/cLIR score matrix between domains, and the sequence viewer
Download the script (.cxc/.pml) and structure file, place them in the same folder, then open the script in ChimeraX or PyMOL

Features

Auto-fetch from AlphaFold Database — PDB structure and PAE matrix fetched directly by UniProt ID
LIS/pLDDT-based domain detection — uses smoothed per-residue pLDDT scores and LIS between adjacent segments to automatically detect domain boundaries, with adjustable parameters:
- Min region length (default: 15 residues)
- Merge LIS threshold (default: 0.35) — adjacent regions with high mutual LIS are merged (higher = less merging)
- pLDDT cutoff (default: 50) — residues below this are treated as disordered linkers
Intramolecular iLIS — calculates iLIS between all detected domain pairs to quantify domain–domain interactions
Score matrix — iLIS (Oranges, lower-left) / cLIR count (white-to-green, upper-right) between domains
PAE/LIS/cLIS maps — same colormaps (bwr, Blues, Greens) with domain boundaries overlaid
Sequence viewer — amino acid letters with LIR (light) and cLIR (dark) highlighting per domain pair
script presets — domain-colored view (default), pLDDT coloring, color by chain, color by polymer
Downloads — .cxc script, .pdb structure, CSV with domain pair metrics

LIR Display Options

All tools include adjustable LIR display settings that affect the visualization script and 3D viewer:

Gap filling (default: 10 residues) — bridges short breaks between LIR regions for continuous cartoon display.
Min segment (default: 3 residues) — removes isolated LIR fragments shorter than the threshold. cLIR residues within removed segments are also removed.

Click Apply after changing values to update the visualization script and 3D viewer.

Sequence Viewer

All four tools include a sequence viewer that displays the actual amino acid sequence (one-letter codes) for each chain or domain. Residues are color-coded based on their interaction status:

A K L

LIR residues (light color)

R D E

cLIR residues (dark color, bold white)

G P S

Non-interacting residues

Residue numbering is shown at intervals. The viewer scrolls horizontally for long sequences.

Color Scheme Reference

Map Colormaps

All tools use consistent, fixed colormaps for maps and matrices. These are not affected by the color presets.

Map	Colormap	Scale
PAE map	blue – white – red (bwr)	0 Å (confident) → 30 Å (uncertain)
LIS map	matplotlib Blues	0 → 1 (higher = stronger interaction)
cLIS map	matplotlib Greens	0 → 1 (higher = stronger contact-filtered interaction)
Score matrix (iLIS)	matplotlib Oranges	Lower-left triangle
Score matrix (cLIR / ipTM)	white → green (or Purples for ipTM)	Upper-right triangle

Visualization Script Colors

The visualization script (.cxc for ChimeraX, .pml for PyMOL) uses chain colors that you can customize via presets. The default is Teal/Coral gradient.

Chain A — LIR (light teal)

Chain A — cLIR (dark teal)

Chain B — LIR (light coral)

Chain B — cLIR (dark coral)

Gradient — light color for LIR, dark color for cLIR (e.g., Blue/Orange, Purple/Gold, Teal/Coral)
Solid — same color for both LIR and cLIR
Named colors — familiar colors (Cornflower/Tomato, SkyBlue/Gold, etc.)
Multi-chain (tab10) — auto-assigned distinct colors per chain, shown for complexes with more than 2 chains
Swap A ↔ B button — reverses chain colors without changing the color combination

Coloring Mode Presets

The Monomer tool provides additional coloring presets beyond the default domain-colored view:

pLDDT coloring — colors residues by AlphaFold confidence: very high (>90), confident (70–90), low (50–70), very low (<50)
color bychain — chain coloring
color bypolymer — polymer coloring

Applying Color Changes

Color presets and custom colors update the visualization script (ChimeraX/PyMOL) automatically when you select a preset. The PAE/LIS/cLIS maps use fixed colorscales and are not affected by color choices.

AFM-LIS Metrics

Metric	Full Name	Definition
iLIS	integrated LIS	√(LIS × cLIS) — geometric mean of LIS and cLIS; balanced score combining confident domain and confident contact
LIS	Local Interaction Score	Average of inversely scaled PAE (0–1, higher is better) within LIA
cLIS	contact-filtered LIS	Average of inversely scaled PAE within cLIA (contact-filtered)
LIR	Local Interaction Residues	Residues in the confident interaction region (PAE ≤ 12 Å)
cLIR	contact-filtered LIR	Residues in direct physical contact (PAE ≤ 12 Å & Cβ ≤ 8 Å)
LIA	Local Interaction Area	Confident interaction area (PAE ≤ 12 Å)
cLIA	contact-filtered LIA	Confident interaction area within contact distance (PAE ≤ 12 Å & Cβ ≤ 8 Å)

Default cutoffs: PAE ≤ 12 Å (confident interaction) and Cβ ≤ 8 Å (direct contact). These can be adjusted before processing in the Universal and Dimer tools.

PAE transformation: For each inter-chain residue pair (i, j), the PAE value is converted to a confidence score: confidence = 1 − (PAE / cutoff) if PAE ≤ cutoff, otherwise 0. This linearly maps PAE = 0 Å to confidence = 1.0 (highest) and PAE = cutoff to 0. The PAE cutoff of 12 Å was determined by ROC analysis to maximize AUC (Kim et al., 2024).

LIS calculation: LIA (Local Interaction Area) is the count of inter-chain residue pairs with PAE ≤ cutoff. LIS is the mean confidence score across all LIA pairs. cLIA further restricts to pairs also in physical contact (Cβ ≤ 8 Å), and cLIS is the mean confidence within cLIA. The inter-chain PAE block is symmetrized by averaging (A→B) and (B→A) directions. iLIS = √(LIS × cLIS), combining interface confidence with direct contact evidence into a single robust metric (Kim et al., 2025).

Contact map: Built from Cβ atom distances (Cα for glycine). Two residues are in contact if their Cβ–Cβ distance ≤ 8 Å. For nucleic acids, phosphorus (P) atoms are used with a 4 Å distance adjustment.

Symmetrization: Results for chain pairs (A→B) and (B→A) are averaged. LIR/cLIR residue sets are the union across all models.

Suggested threshold: iLIS ≥ 0.223 was established at 10% FDR using large-scale Y2H reference datasets in yeast (Yu et al., 2008), fly (Tang et al., 2023), and human (Braun et al., 2009). This threshold was determined using AlphaFold-Multimer predictions (via ColabFold). See Kim et al., 2025 supplementary text for details.

Caution: Computational predictions should be interpreted with care. The iLIS ≥ 0.223 threshold was derived from AlphaFold-Multimer (ColabFold) predictions with specific reference datasets. Other prediction platforms (AlphaFold 3, Boltz, Chai-1, OpenFold) may require different cutoffs due to differences in model architecture, training data, and confidence calibration. Each platform produces distinct PAE distributions and ipTM scales, which can affect iLIS values. Optimal thresholds may also vary depending on the species, protein types, and experimental system. Experimental validation is highly recommended before drawing biological conclusions from predicted interactions.

Batch Analysis

For large-scale batch analysis of many predictions, use lis.py from the AFM-LIS repository. It supports all the same platforms (AlphaFold3, ColabFold, Boltz, Chai-1, OpenFold3), auto-detects the prediction format, and outputs CSV.

                python lis.py /path/to/predictions/

                python lis.py /path/to/predictions/ -w 4  # parallel with 4 CPUs

Features: .gz/.xz decompression, incremental CSV output (safe to interrupt and resume), progress bar with ETA.

References

AFM-LIS

Kim et al. 2024 — Enhanced Protein-Protein Interaction Discovery via AlphaFold-Multimer
Kim et al. 2025 — A Structure-Guided Kinase–Transcription Factor Interactome Atlas Reveals Docking Landscapes of the Kinome
AFM-LIS GitHub — Code and documentation for calculating integrated Local Interaction Score (iLIS)

Structure Prediction

Jumper et al. 2021 — Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589.
Evans et al. 2022 — Protein complex prediction with AlphaFold-Multimer. bioRxiv.
Abramson et al. 2024 — Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630, 493–500.
Mirdita et al. 2022 — ColabFold: making protein folding accessible to all. Nature Methods 19, 679–682.

Confidence Metrics

Dunbrack 2025 — Rēs ipSAE loquuntur: What’s wrong with AlphaFold’s ipTM score and how to fix it. bioRxiv.

Y2H Reference Datasets (iLIS Benchmarks)

Yu et al. 2008 — High-quality binary protein interaction map of the yeast interactome network. Science 322, 104–110.
Tang et al. 2023 — Next-generation large-scale binary protein interaction network for Drosophila melanogaster. Nature Communications 14, 2177.
Braun et al. 2009 — An experimentally derived confidence score for binary protein-protein interactions. Nature Methods 6, 83–90.

Visualization

Sehnal et al. 2021 — Mol* Viewer: modern web app for 3D visualization and analysis of large biomolecular structures. Nucleic Acids Research 49, W431–W437.
Pettersen et al. 2021 — UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Science 30, 70–82.
The PyMOL Molecular Graphics System, Schrödinger, LLC. pymol.org

Databases

UniProt Consortium 2023 — UniProt: the Universal Protein Knowledgebase in 2023. Nucleic Acids Research 51, D523–D531.
Varadi et al. 2022 — AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Research 50, D439–D444.
Hu et al. 2025 — FlyRNAi.org 2025 update: expanded resources for new technologies and species. Nucleic Acids Research 53(D1), D958–D965.
Hu et al. 2018 — Molecular Interaction Search Tool (MIST): an integrated resource for mining gene and protein interaction data. Nucleic Acids Research 46(D1), D567–D574.
Paysan-Lafosse et al. 2023 — InterPro in 2022. Nucleic Acids Research 51(D1), D418–D427.
Ozturk-Colak et al. 2024 — FlyBase: updates to the Drosophila genes and genomes database. Genetics 227(1), iyad211.

This Tool

LIVIA GitHub — Source code for all visualization and analysis tools

Cite this tool

If you use iLIS metric in your research, please cite:

                    Kim, A.-R. et al. (2025). A Structure-Guided Kinase–Transcription Factor Interactome Atlas Reveals Docking Landscapes of the Kinome. bioRxiv. https://doi.org/10.1101/2025.10.10.681672

                    Kim, A.-R. et al. (2024). Enhanced Protein-Protein Interaction Discovery via AlphaFold-Multimer. bioRxiv. https://doi.org/10.1101/2024.02.19.580970

If you use the FlyPredictome data, please also cite:

                    Hu, Y. et al. (2025). FlyRNAi.org 2025 update — expanded resources for new technologies and species. Nucleic Acids Res. 53(D1), D958–D965. doi:10.1093/nar/gkae917
                