Gradient Biotech

Sample data

If you don't have your own cohort data yet, use these minimal examples to try each pipeline before uploading real study data. Save any of the tables below as a .csv file.

Mutation landscape

A minimal MAF-style mutation file:

Hugo_Symbol,Tumor_Sample_Barcode,Variant_Classification,Chromosome,Start_Position,Reference_Allele,Tumor_Seq_Allele2
TP53,S1,Missense_Mutation,17,7579472,C,T
KRAS,S1,Missense_Mutation,12,25398284,G,A
TP53,S2,Nonsense_Mutation,17,7578406,C,A
PIK3CA,S2,Frame_Shift_Del,3,178936091,T,TA
BRCA1,S3,Missense_Mutation,17,43071077,C,G
EGFR,S3,Silent,7,55249071,A,G
CDKN2A,S3,Splice_Site,9,21971120,G,T

Optional matching copy number file:

sample_id,gene,chromosome,alteration
S1,EGFR,7,amplification
S2,PTEN,10,deletion
S3,CDKN2A,9,deletion

With this data, expect: 3 samples, 7 variants, TMB per megabase greater than 0 for all samples, TP53 as the top mutated gene, and a dominant mutational signature.

Cell-cell communication

Per-cell expression:

cell_id,TGFB1,TGFBR1,CXCL9,CXCR3,CD274,PDCD1
tumor_1,5.0,1.0,0.0,1.0,6.0,0.0
tumor_2,4.0,1.0,0.0,1.0,5.0,0.0
t_1,0.0,4.0,3.0,5.0,0.0,5.0
t_2,0.0,4.0,4.0,5.0,0.0,4.0
myeloid_1,2.0,2.0,1.0,0.0,1.0,0.0
myeloid_2,1.0,2.0,1.0,0.0,1.0,0.0

Matching metadata:

cell_id,cell_type,condition
tumor_1,Tumor,baseline
tumor_2,Tumor,treated
t_1,T cell,baseline
t_2,T cell,treated
myeloid_1,Myeloid,baseline
myeloid_2,Myeloid,treated

With this data, expect a handful of ligand-receptor pairs tested across 3 cell types, with condition-comparison deltas between baseline and treated.

Survival analysis

Clinical outcomes:

patient_id,time_to_event_days,event_observed,treatment_arm
P1,900,0,anti-pd1
P2,820,0,anti-pd1
P3,760,1,anti-pd1
P4,320,1,chemo
P5,280,1,chemo
P6,240,1,chemo

Optional molecular features for Cox covariates:

patient_id,tmb,cd8_score
P1,9.0,2.0
P2,8.0,1.8
P3,7.5,1.6
P4,2.0,-0.8
P5,1.0,-1.0
P6,1.5,-0.9

With this data, expect 6 patients split into 2 groups when stratified by treatment arm, with Kaplan-Meier curves and Cox regression output.

Suggested first walkthrough

  1. Create an oncology study
  2. Upload the mutation file above and run Mutation landscape
  3. Upload the clinical file above and run Survival analysis, stratifying by treatment arm
  4. Run Interpret, selecting both completed runs
  5. Review results on the Mutations and Survival pages, and in run history

See Quick start for the full walkthrough, and File formats for exact column requirements per pipeline.