Interactive dashboard for exploring 182,362 complete plasmids from PLSDB 2025 (72,556) + NCBI Nucleotide (109,806 additional), powered by a local SQLite database. No API rate limits, instant startup.

Features

Overview

Topology, source, kingdom distributions, temporal trends, GC content, and plasmid length distributions across the entire database.

Taxonomy Explorer

Top 20 host genera with interactive search by genus or species.

AMR Analysis

163 antimicrobial resistance drug classes and individual gene frequencies coloured by their real drug class from the database.

Plasmid Viewer

Interactive circular plasmid map with zoom and hover tooltips. Enter any NCBI accession to visualize:

• CDS genes (forward/reverse strand) with product descriptions
• AMR genes highlighted in red (blaKPC-2, blaTEM, etc.)
• MOB elements (mobilization genes: mobA, traI, etc.)
• Toxin-Antitoxin systems (relE, higA, ccdA/B, vapB, etc.)
• Virulence factors (iutA, sitA, hlyA, etc.)
• QAC resistance (qacE, smr, etc.)
• GC content deviation ring

Works for both PLSDB plasmids and any NCBI GenBank accession (automatic fallback).

Inc Groups & Mobility

Incompatibility group (replicon type) distribution, predicted mobility, relaxase types, and MPF types.

Correlations (n = 72,556)

Cross-feature co-occurrence analysis using the entire database:

1. AMR Drug Class vs Inc Groups & Mobility -- heatmap and grouped bar
2. Heavy Metal Resistance -- mercury, arsenic, copper, silver, tellurium genes and Inc group associations
3. Phage & Mobile Elements -- distribution and mobility correlation
4. pMLST Distribution -- scheme frequency, alleles, mobility
5. Virulence Factors (6,145 plasmids) -- top genes, Inc group and mobility
6. Toxin-Antitoxin Systems (40,759 plasmids) -- vapB, ccdA/B, pemI/K from PGAP annotations
7. QAC Resistance (7,023 plasmids) -- quaternary ammonium compound resistance vs Inc groups

Geography (57,751 geolocated plasmids)

1. Global Plasmid Map -- scatter map coloured by mobility, hover for accession/country/Inc group
2. Country Comparison -- mobility and Inc group distribution by top 20 countries
3. Temporal Spread Animations -- animated timeline (2000-2024) showing cumulative plasmid reports per country; separate animations for specific Inc groups (IncFIB, IncFII, etc.) and mobility types (conjugative, mobilizable, non-mobilizable)
4. Plasmid Host & Source Correlations -- plasmid distribution by host category (Human 11,464 / Animal 5,925 / Soil 1,576 / Water 1,332 / Food 1,059 / Environment 959), with heatmaps for Inc groups and AMR drug classes by host

Plasmid Lookup

Search any accession to view metadata: length, GC, topology, taxonomy, source, creation date.

Quick Start

pip install -r requirements.txt
python build_db.py    # downloads PLSDB CSVs from Figshare, builds SQLite DB
python app.py         # open http://localhost:8050

Database

The dashboard uses a local SQLite database (data/plsdb.db, ~200 MB) built from PLSDB Figshare CSV exports including PGAP protein annotations. This eliminates all API calls and enables full-database analysis.

Tables: nuccore (72,556), taxonomy (7,439), typing (72,556), amr (251,138), plasmidfinder (60,619), typing_markers (333,648), pgap_features (613,902), biosample_location (23,522).

Deploy to Heroku

heroku create plasmidnet
git push heroku main
heroku open

The bin/post_compile hook automatically downloads PLSDB CSVs from Figshare and builds the SQLite DB during deploy. No large files in git.

Deploy to Render (free tier)

1. Connect your GitHub repo at render.com
2. Create a new Web Service
3. Set build command: pip install -r requirements.txt && bash bin/post_compile
4. Set start command: gunicorn app:server --bind 0.0.0.0:$PORT

Tech Stack

• Dash + Plotly for interactive visualizations
• SQLite for local full-database queries (72,556 plasmids, 251K AMR, 614K PGAP annotations)
• PLSDB 2025 as data source (version 2024_05_31_v2)
• NCBI GenBank as fallback for plasmid viewer (CDS + GC content)
• Gunicorn for production serving