Ask TCGA a cancer-genomics question.

Insight™ picks the cohort and runs the analysis.

Insight came first and is still in active development. DrugAdopt is built on it.

Try Insight

Ask from your phone.

Insight keeps the cohort and analysis when you continue on desktop.

9:415G ▰

Insight BRCA

Where are TP53 mutations concentrated in breast cancer?

Cohort matched · analysis checked

TP53 coding mutation hotspots

R175HR273HR196*R213*Y220C

1,066 gene-profiled tumors

Caveat Rates are per profiled tumor, not TP53 prevalence.

Ask a follow-up… ↑

BRCA · TP53 differential expression

Analysis recipe + provenance

analysis_recipe.py

# Exported analysis recipe
study = "brca_tcga_pan_can_atlas_2018"
assays = ["mutation", "rna_seq"]
contrast = "TP53_mutant - TP53_wild_type"

matched = match_samples(study, assays)
model = differential_expression(
    matched,
    design="~ TP53_status + PAM50_subtype"
)

model.volcano(
    fdr=0.05,
    abs_log2fc=1.0
)

Dataset: TCGA PanCancer Atlas · BRCA
Sample match: Mutation + RNA-seq
Contrast: TP53-mutant vs wild-type
Adjustment: Benjamini–Hochberg FDR
Unmodeled: Tumor purity

The recipe adjusts for PAM50 subtype. Tumor purity is not modeled.

FOCUS BRCA TCGA PanCanAtlas TP53

PROFILED: N=1,066

✓

What I would check next

Ask a follow-up question…↑

#1 TP53 coding mutation hotspots in TCGA breast cancer protein position · profiled tumors

21100 samples

R175H

R196*

R213*

Y220C

R273H

TAD

PRD

DNA-binding

Tet.

CTD

Five of 50 ranked changes shown 1,066 profiled 50 ranked changes

#2 Differential expression by TP53 status in TCGA breast cancer matched mutation + RNA-seq samples

Illustrative genes and positions · not TCGA estimates FDR 0.05 |log2FC| ≥ 1

One BRCA/TP53 question: asked on a phone, analyzed and followed up on desktop, then exported as a notebook recipe.