Brown Biotech Biostatx
Biostatistics and bioinformatics for decision-ready outputs — same-week turnaround.
Get in touchAbout us
Biostatx is Brown Biotech's statistical-analysis lane: regression, survival, dose-response, biomarker discovery, and reproducible figure-grade outputs. We pair R/Python tooling with a System-of-Intelligence reasoning layer that lets clients ask 'is this difference real, and what does it mean for the next experiment?' — not just 'here is a p-value'. Live RAG preview wired to PRISM FastAPI; full report delivered within a week for typical scopes.
What we offer
Regression & survival analysis
Cox proportional hazards, Kaplan-Meier with competing risks, mixed-effects models. Output: publication-grade figures + interpretation memo.
Dose-response and PK/PD modeling
Non-linear fits, Emax/Emax-Hill models, basic PK/PD parameter estimation. Used in early-stage in-vivo / in-vitro screens.
Biomarker discovery
Multi-omics feature selection (DEG, methylation, proteomics), pathway enrichment, cross-cohort validation. Output: ranked candidate table + biological interpretation.
Reproducible figure pipeline
R Markdown + ggplot2 + RO-Crate provenance. Every figure ships with the script, data hash, and environment snapshot.
PRISM RAG live preview
Visitors can ask research questions and get real answers from a 19-chunk PubMed-backed RAG before submitting the formal inquiry.
Frequently asked questions
What's the typical turnaround?
Standard scope: 5 business days from data receipt. Complex multi-omics analyses: 2–3 weeks. Rush service available for 3-day delivery with surcharge.
What data formats do you accept?
CSV, TSV, RDS, Parquet, Excel. For omics: raw count matrices, processed Seurat/Scanpy objects, GEO supplements. We sign a data-use agreement before transfer.
Do you sign publication co-authorship?
If our analysts contribute substantive statistical design, yes. For routine analysis, we deliver an analytical report but reserve the right to be acknowledged but not co-authored.
Can you handle clinical trial data?
Yes — under CDA. We follow CONSORT and STROBE reporting guidelines. Sample-size re-estimation and adaptive-design support available.
How is reproducibility ensured?
Every output script runs from a pinned environment (renv or conda env yml), data is hashed (sha256), and RO-Crate metadata captures inputs, code, and outputs.