DILS

DILS means Demographic Inferences with Linked Selection.

DILS is a workflow for demographic inference from multilocus DNA sequence data. It uses Approximate Bayesian Computation (ABC) for model comparison, parameter estimation, and goodness-of-fit analyses.

Documentation

• User manual: docs/user_manual.md
• Cluster/admin deployment guide: docs/cluster_deployment.md
• Streamlit configuration example: streamlit/settings.example.yaml

Ways to Use DILS

DILS can currently be used in two main ways:

1. Streamlit web interface: upload a FASTA file, configure an analysis, submit it through Slurm, download the results archive, and inspect outputs in the Results viewer.
2. Command-line Snakemake workflow: run the DILS Snakefiles directly with a YAML configuration file.

The repository may still contain the older R/Shiny web interface under webinterface/. It is kept as legacy code and is not the recommended interface for new deployments.

Streamlit Web Interface

The Streamlit interface lets users:

• submit a DILS analysis;
• monitor Slurm job status;
• download the final .tar.gz results archive;
• inspect results in the Results viewer;
• read the integrated Help page.

Local launch example:

streamlit run streamlit/app.py

For configuration, copy and edit:

cp streamlit/settings.example.yaml streamlit/settings.yaml

streamlit/settings.yaml is local configuration and is ignored by Git. Cluster deployment is documented in docs/cluster_deployment.md.

Example Installation: Streamlit Interface

Clone the repository:

git clone https://github.com/popgenomics/DILS.git
cd DILS

Create and activate a Python environment, then install the Streamlit requirements:

python3 -m venv .venv_streamlit
source .venv_streamlit/bin/activate
python -m pip install --upgrade pip
python -m pip install -r streamlit/requirements.txt

Create a local Streamlit settings file:

cp streamlit/settings.example.yaml streamlit/settings.yaml

For local testing, set the runs directory to streamlit_runs:

sed -i 's|runs_root: /path/to/dils_runs|runs_root: streamlit_runs|' streamlit/settings.yaml

Check the configured DILS paths:

grep -n "dils_bin_dir\|snakefile_1pop\|snakefile_2pop" streamlit/settings.yaml

Expected output:

dils_bin_dir: bin
snakefile_1pop: bin/Snakefile_1pop
snakefile_2pop: bin/Snakefile_2pop

Check the Streamlit upload limit:

cat .streamlit/config.toml

Expected output:

[server]
maxUploadSize = 10240

Launch the interface:

streamlit run streamlit/app.py

Notes:

• streamlit/settings.yaml is local and ignored by Git.
• runs_root: streamlit_runs is for local testing only.
• For cluster deployment, use docs/cluster_deployment.md.
• If no bundled example archive is found under example/, use “Upload .tar.gz” in the Results viewer.

Command-Line Snakemake Workflow

DILS can also be run directly with Snakemake and a YAML configuration file.

1-population analysis:

snakemake --snakefile bin/Snakefile_1pop -p -j 10 --configfile config_1pop.yaml

2-population analysis:

snakemake --snakefile bin/Snakefile_2pop -p -j 10 --configfile config_2pop.yaml

The Streamlit interface generates YAML internally. Command-line users should provide their own YAML file.

YAML Configuration Examples

1-Population Example

mail_address: user@example.org
infile: /path/to/input.fas
region: coding
nspecies: 1
nameA: PopulationA
nameOutgroup: NA
lightMode: TRUE
config_yaml: /path/to/config_1pop.yaml
timeStamp: my_1pop_run
max_N_tolerated: 0.2
Lmin: 100
nMin: 6
mu: 0.00000002763
rho_over_theta: 0.5
N_min: 1000
N_max: 500000
Tchanges_min: 100
Tchanges_max: 1000000

2-Population Example

mail_address: user@example.org
infile: /path/to/input.fas
region: coding
nspecies: 2
nameA: PopulationA
nameB: PopulationB
nameOutgroup: NA
lightMode: TRUE
useSFS: 0
config_yaml: /path/to/config_2pop.yaml
timeStamp: my_2pop_run
population_growth: constant
modeBarrier: bimodal
max_N_tolerated: 0.2
Lmin: 100
nMin: 6
mu: 0.00000002763
rho_over_theta: 0.5
N_min: 1000
N_max: 500000
Tsplit_min: 10000
Tsplit_max: 1750000
M_min: 1
M_max: 40

Runtime Dependencies

Runtime dependencies depend on whether DILS is used through the Streamlit interface or directly with Snakemake. In general, DILS requires:

• Python;
• PyPy where required by legacy scripts;
• R and required R packages;
• Snakemake;
• Slurm for cluster submission;
• Streamlit dependencies from streamlit/requirements.txt;
• C helper binaries such as msnsam and RNAseqFGT.

Cluster administrators should use docs/cluster_deployment.md for deployment-specific details.

Repository Layout

• bin/: DILS Snakemake workflows and analysis scripts.
• streamlit/: current Streamlit web interface.
• docs/: user and administrator documentation.
• example/: example input and results files.
• webinterface/: legacy R/Shiny interface.
• msnsam/, RNAseqFGT_src/: C sources and helper binaries.

Citation / Support

For scientific background, usage details, and references, see the user manual:

docs/user_manual.md