- R 79.4%
- Python 12.9%
- Shell 4.7%
- Jupyter Notebook 2.9%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
|
|
||
| analysis | ||
| figures | ||
| legacy_codes | ||
| preprocessing_pipeline | ||
| tables/R | ||
| .gitattributes | ||
| .gitignore | ||
| LICENSE | ||
| README.md | ||
| SETUP.md | ||
The GLASS consortium
Overview
The Glioma Longitudinal AnalySiS (GLASS) consortium consists of clinicians, bioinformaticians, and basic science researchers from leading institutions across the world striving to better understand glioma tumor evolution and to expose its therapeutic vulnerabilities. The code in this respository was used to generate the figures and perform the analyses described in our 2025 publication in BioRxiv.
R code used to make each of the figures can be found in the figures/R subdirectory.
Data Release version 2025.
The data analyzed using this code was part of the fourth release of the GLASS dataset which was made available in 2025. This dataset is managed internally using the PostgreSQL database management system. These data are under active curation so future versions will include additional data as well as correct potential errors.
Data Download
The GLASS data can be downloaded from the Tables page here and the Files page here. It is also possible to query the data directly using the the API by using queries. You can read more about that here.
Prior releases
Previous projects and associated repositories are listed below:
Barthel FP, Johnson KC, Varn FS, et al. Longitudinal molecular trajectories of diffuse glioma in adults. Nature. 2019; 576(7785): 112-120. Paper. Code.
Varn FS, Johnson KC, Martinek J, et al. Glioma progression is shaped by genetic evolution and microenvironment interactions. Cell. 2022; 185(12): 2184-2199.e16. Paper. Code.
Repository layout
The repository separates the code behind this manuscript from the standardized GLASS infrastructure it was built on top of. The first three directories below are the former; the last two are the latter.
| Directory | What it holds |
|---|---|
analysis/ |
Analyses that generate derived results such as copy number, structural variants, mutational signatures |
figures/ |
One script per manuscript figure panel |
tables/ |
One script per manuscript table, mostly cohort description and survival models |
legacy_codes/ |
Code from the prior GLASS publications listed above, retained for provenance |
preprocessing_pipeline/ |
The GLASS workflow that produced the database itself |
Environment and database configuration for analysis/, figures/ and tables/ is documented separately in SETUP.md.
analysis/
Derived analyses: everything that computes a result which the figures and tables then read. 23 files in six subdirectories, split by language and by role — R and Python for code you call directly, bash/ and slurm/ for the drivers that orchestrate multi-stage runs, snakemake/ for the cluster workflows.
| Subdirectory | Files | Contents |
|---|---|---|
analysis/R/ |
12 | Chromothripsis calling (ShatterSeek), ecDNA/amplicon analysis (AmpliconSuite), allele-specific copy number (ASCAT), whole-genome doubling, fraction-of-genome-altered calculation, GISTIC input preparation, copy-number signature exploration, telomere-length normalisation |
analysis/SQL/ |
4 | Query definitions consumed by the scripts and pipelines — CNA extraction, the hypermutant and non-hypermutant dN/dS inputs, GISTIC preparation |
analysis/bash/ |
2 | Two self-contained, resume-aware pipeline drivers: mutational_signature_pipeline.sh (repeat-region filtering → SigProfilerMatrixGenerator/Extractor → Palimpsest deconvolution) and kataegis_pipeline.sh (clustered-mutation / kataegis calling via SigProfilerSimulator + SigProfilerClusters). Both accept --dry-run |
analysis/python/ |
1 | ampsuite_manifest.py, the AmpliconSuite sample-manifest builder called by ampsuite.smk |
analysis/slurm/ |
1 | GLASS-I_SvABA.sbatch, the SvABA structural-variant calling job |
analysis/snakemake/ |
3 | Cluster workflows over BAM files: ampsuite.smk, ascat.smk, sigprofilertoolkit.smk. See the note in SETUP.md §4.3 — their helper-script paths need editing before they run |
figures/
One script per figure panel, 28 files. Each is standalone: it opens the GLASS database, builds its own plotting table, and writes the panel.
| Subdirectory | Files | Contents |
|---|---|---|
figures/R/ |
27 | Cohort description and QC Sankey; swimmer's plot; copy-number heatmap and per-subtype CN segment plots (astrocytoma and oligodendroglioma, primary / recurrence / delta); fraction-of-genome-altered, with and without hypermutants and by grade; chromosomal-instability landscape and CIN markers; IDH mutation figures; mutation burden; mutational-signature boxplots; dN/dS selection; sSNV mutual exclusivity; TMZ cycles; and Kaplan–Meier curves for CNA, drivers (GENIE), hypermutation, ID8 and sSNV burden |
figures/python/ |
1 | The world map of cohort contributing centres |
tables/
One script per manuscript table, 16 files, all R.
| Group | Contents |
|---|---|
| Cohort description | Baseline table, molecular cohort summary, prior-treatment table |
| Drivers | Driver gene table, drivers by prior treatment, IDH mutation table, McNemar test on paired CNA changes |
| Survival | Univariable and multivariable models for three endpoints — overall survival (OS), post-recurrence survival (PRS) and time to recurrence (TTR) — each split into clinical and molecular covariates |
Output is formatted with gt and gtsummary.
legacy_codes/
Created separately, outside the scope of this manuscript's analysis code, under the same standardized GLASS pipeline protocols. 333 files of R (308), Python (19) and Julia (5) from the earlier GLASS publications listed under Prior releases above, retained for provenance and reproducibility of those papers.
Organised by topic under legacy_codes/R/: single-cell deconvolution (CIBERSORTx), neoantigen analysis, SNV and CNV processing, manifest handling, mutation timing, figures and tables for the 2019 and 2022 papers, and a Shiny app. The Julia code covers subclonal selection analysis.
This directory is not maintained against the current data release and its dependencies are not part of SETUP.md.
preprocessing_pipeline/
Created separately, through the standardized GLASS consortium pipeline protocols described in the 2025 publication linked above and established in the prior GLASS releases listed under Prior releases. 251 files. This is the workflow that produced the PostgreSQL database the rest of the repository queries — it is not re-run as part of reproducing a figure, and it is not covered by SETUP.md.
| Subdirectory | Files | Contents |
|---|---|---|
sql/ |
125 | The schema and queries that populate and serve the database, grouped by domain: clinical, variants, snv, cnv, drivers, mut_freq, mut_sig, pyclone, timing, neutrality, neoag, immune, lohhla, expression, heatmap, metadata, set, dndscv, figures |
snakemake/ |
30 | Per-stage workflow modules, driven by the top-level Snakefile |
envs/ |
44 | One conda environment specification per tool — alignment, GATK4, Sequenza, PyClone, Manta, Delly, LUMPY, OptiType, pVACseq, MiXCR, Kallisto, and others |
conf/ |
— | Cluster and pipeline configuration |
bin/, jar/ |
— | Bundled executables and Java dependencies |
dag/, dbm/ |
— | Workflow DAGs and the database schema diagram |
Because this directory carries its own environment specifications in envs/, use those rather than the instructions in SETUP.md if you need to re-run any part of it.
Where to start
- Reproducing a figure or table → read
SETUP.md, configure the database connection, then run the single script for that panel. - Reproducing a derived analysis → the corresponding script in
analysis/R/, or one of the two drivers inanalysis/bash/, which accept--dry-run. - Understanding how the data was generated →
preprocessing_pipeline/and the methods of the manuscript. - Looking for code from the 2019 or 2022 papers →
legacy_codes/, or the dedicated repositories linked under Prior releases.
