≡
NOSE Pipeline Overview
From assembled genomes to novel species · quality → novelty → placement → prevalence → function → metabolism
NOSE — Novel Species Identification Pipeline
An automated, reproducible workflow that takes assembled genomes and identifies and characterizes potential novel species across six sequential/branching steps. Developed at the Computational Systems Biology Lab, IIT Madras.
Manual setup — run once on your server
You're viewing this dashboard right now, which already proves steps 1, 2
and 4 are done on this machine — they're shown collapsed below for reference (e.g. setting up
another server), not as something left to do. Only step 3 is checked live since the
dashboard itself doesn't need conda, but every module does.
✓
Get the code onto the server Already done
Show command
git clone https://github.com/RamanLab/NOSE.git && cd NOSE
✓
Create the dashboard's Python environment Already done
Show command
python3 -m venv .venv && source .venv/bin/activate && pip install -r requirements.txt
3
Make sure conda/mamba is available — each module pulls its own tool environment on first run
Checking…
Show command
conda --version
✓
Launch the dashboard (reachable from other machines on your network) Already done — you're here
Show command
python Run-pipeline.py --host 0.0.0.0 --port 5050
From install to result, in one line: once the dashboard is running, open it in a browser,
fill each module's genome/database paths under its Config tab (or use the sidebar's
"Genome Source" + Apply to All Modules), then either run modules one at a time or click
Run Full Pipeline. The diagram below shows exactly what happens next — M1 checks
quality, M2 decides which genomes are novel, and that novelty gate is what feeds M3, M5 and M6.
Live progress and finished results for each run show up further down under Module Status.
Credentials & Licenses
A quick note before you fill in Config: every key and license below is yours NOSE doesn't ship or share any of them.
This dashboard runs locally on your machine nothing leaves it except calls to the services you configure below.
N
NCBI API Key & Email
Required · Free
Resolves real organism names from GCA/GCF accessions. Grab a free key at
ncbi.nlm.nih.gov/account/settings
→ API Key Management, then drop it into the sidebar's Credentials section it applies to all 6 modules.
C
CPLEX (Module 6 solver)
Your own license
Point
CPLEX Library Path in Module 6's Config — or the same field on the
Full Pipeline Run page under M6 · CarveMe Settings — at your own licensed installation
(e.g. via the IBM Academic Initiative).
NOSE never bundles or redistributes CPLEX itself.
D
Reference databases
Free & public
GTDB-Tk, CheckM2, antiSMASH, COG/CDD, geNomad, BLAST 16S download to your own server via each
module's Config tab. You'll always be asked to confirm before a download starts. Not bundled with NOSE.
Keep it to your own accounts nothing above is logged or shared beyond the service it's configured for (
nose_config.yaml and every Module*/config.yaml are git-ignored by default).How it works — pipeline workflow
Click any step to open its config
Idle
Running
Done
Failed
Dashed box = novelty gate (ANI < 95% only)
📊
Module Status
Live results and run status for each module — no workflow explanation here, see Overview for that
-
Total genomes (M2)
-
Novel · ANI < 95%
-
Known species
0/6
Modules completed
Per-module report
Loading module status…
▶
Full Pipeline Run
Fill in every step's settings below, save once, then run all six steps end to end
▶ Pipeline
Pipeline configuration — all steps
▼ End of config