Imputation · Full report · powered by Responder Atlas

Indication Full.

The paid per-compound report — a deeper application of Responder Atlas. Everything Indication Lite gives you, plus the two things a triage decision usually needs next: which subpopulations concentrate the response, and which biomarkers discriminate them.

See the full sample report Commission a report

Included

Responder subpopulations

Cell-line groups whose predicted response passes both strength and coherence cutoffs, with tissue composition, node identity, and margins visible.

Included

Candidate biomarkers

Ranked expression and mutation features that discriminate predicted responders from background — with effect size, cross-validated importance, and nominal p, plus an underpowered flag when n is thin.

Included

Ranked indication signal

The same 17-tissue signal Indication Lite returns, in server order, kept alongside the paid layers so provenance stays legible.

Included

Confidence tier and provenance

Tanimoto-neighbourhood confidence, nearest training-set drug, and basis (measured vs. imputed) surfaced on every report.

Hypothesis-generating, not validated. The Full Responder Report is a triage layer built on cell-line data — leads to test, not clinical decisions. Every report carries the confidence tier and disclaimer that governs its interpretation.

Sample excerpt

A preview of the paid report.

The excerpt below is trimmed from the committed gefitinib illustrative sample — the same rendering the live sample uses, with a subset of the responder groups and biomarker rows.

Product · Full Responder Report

gefitinib

Basis: SMILES imputedModel: v0.3

HYPOTHESIS-GENERATING, NOT VALIDATED PREDICTION

SMILES: COc1cc2ncnc(Nc3ccc(F)c(Cl)c3)c2cc1OCCCN1CCOCC1

Hypothesis

HYPOTHESIS: responder structure concentrated in 18 cell-line group(s) (148 cells), led by uterus. Candidate biomarkers below are LEADS TO TEST, not validated markers.

Ranked indications

6 indications · server-ranked

01Esophaguselevated
Predicted viability change+44.2%
reduces viability — predicted cell killing
Selectivity vs. this drug's average+62%
more responsive than this drug's own average
02Upper Aerodigestiveelevated
Predicted viability change+32.6%
reduces viability — predicted cell killing
Selectivity vs. this drug's average+34%
more responsive than this drug's own average
03Urinary Tractelevated
Predicted viability change+30.5%
reduces viability — predicted cell killing
Selectivity vs. this drug's average+30%
more responsive than this drug's own average
04Pancreaselevated
Predicted viability change+18.7%
reduces viability — predicted cell killing
Selectivity vs. this drug's average+11%
more responsive than this drug's own average
05Uteruselevated
Predicted viability change+17.5%
reduces viability — predicted cell killing
Selectivity vs. this drug's average+10%
more responsive than this drug's own average
06Ovaryelevated
Predicted viability change+14.8%
reduces viability — predicted cell killing
Selectivity vs. this drug's average+6%
more responsive than this drug's own average

Two different things. Predicted viability change is the absolute effect — whether the drug is predicted to reduce viability (kill; bar right of zero) or increase it (proliferate; bar left). Selectivity is how that indication compares to this drug's own average across all indications (bar right = more responsive). Both bars diverge around a zero line.

Excerpt · showing 6 of 17 ranked indications.

Responder subpopulations

Cell-line groups whose predicted response passes both strength and coherence cutoffs. Hypothesis-generating, not clinical.

4 groups · server-ranked

Flagged groups: 18
Flagged cells: 148
Nodes considered: 485
Strength cutoff (log₂): -0.2065
Coherence cutoff: 0.1934
Indications covered: Bile DuctBreastLungOvaryPancreasUrinary TractUterus

d0_h0sr#0
strongcoherent
Pancreas
13 cells · geometry d0_h0sr, node 0
- Pancreas4
- Urinary Tract4
- Bile Duct1
- Lung1
- Upper Aerodigestive1
- Uterus1
- Breast1
Score (log₂)
-0.356
Coherence
0.151
Δ vs cut
+0.150
d3_h1sr#7
strongcoherent
Bile Duct
5 cells · geometry d3_h1sr, node 7
- Bile Duct1
- Upper Aerodigestive1
- Gastric1
- Pancreas1
- Ovary1
Score (log₂)
-0.321
Coherence
0.156
Δ vs cut
+0.115
d10_h1sr#218
strongcoherent
Urinary Tract
11 cells · geometry d10_h1sr, node 218
- Urinary Tract4
- Uterus2
- Bile Duct1
- Pancreas1
- Kidney1
- Lung1
- Liver1
Score (log₂)
-0.294
Coherence
0.190
Δ vs cut
+0.088
d3_h0sr#3
strongcoherent
Uterus
21 cells · geometry d3_h0sr, node 3
- Uterus3
- Pancreas2
- Upper Aerodigestive2
- Ovary2
- Liver2
- Urinary Tract2
- Bile Duct1
- Gastric1
- Lung1
- Colorectal1
- Central Nervous System1
- Mesothelioma1
- Esophagus1
- Soft Tissue1
Score (log₂)
-0.270
Coherence
0.187
Δ vs cut
+0.063

Excerpt · showing 4 of 18 responder groups.

Candidate biomarkers — LEADS TO TEST, not validated markers

Framing from the schema: CANDIDATES TO TEST — not validated biomarkers (imputed transfer is weak)

Expression CV-AUC: 0.760
Mutation CV-AUC: 0.447
Cells (responders / background): 148 / 132

Expression

SOS1in CART
low expr → stronger response
RF imp.
0.0165
CV imp.
0.0145 ±0.0036
Nominal p
1.65e-7
Cohen's d
-0.548
MYCLin CART
low expr → stronger response
RF imp.
0.0102
CV imp.
0.0090 ±0.0009
Nominal p
6.97e-6
Cohen's d
-0.583
TSC2
low expr → stronger response
RF imp.
0.0099
CV imp.
0.0073 ±0.0035
Nominal p
0.0026
Cohen's d
-0.311
INSR
low expr → stronger response
RF imp.
0.0092
CV imp.
0.0068 ±0.0026
Nominal p
3.74e-6
Cohen's d
-0.592
CRKL
low expr → stronger response
RF imp.
0.0082
CV imp.
0.0061 ±0.0022
Nominal p
8.56e-4
Cohen's d
-0.357
TAP2
high expr → stronger response
RF imp.
0.0073
CV imp.
0.0070 ±0.0016
Nominal p
1.30e-4
Cohen's d
0.409

Mutation

EP300in CART
mut present → stronger response
RF imp.
0.0131
CV imp.
0.0124 ±0.0014
Nominal p
0.0118
Risk diff.
0.116
KRAS
mut present → stronger response
RF imp.
0.0130
CV imp.
0.0123 ±0.0009
Nominal p
1.05e-1
Risk diff.
0.091
HLA-A
mut absent → stronger response
RF imp.
0.0121
CV imp.
0.0114 ±0.0041
Nominal p
1.05e-1
Risk diff.
-0.061
PIK3CA
mut absent → stronger response
RF imp.
0.0091
CV imp.
0.0092 ±0.0019
Nominal p
3.75e-1
Risk diff.
-0.045
TP53
mut present → stronger response
RF imp.
0.0086
CV imp.
0.0091 ±0.0006
Nominal p
6.76e-1
Risk diff.
0.028
SETD2
mut absent → stronger response
RF imp.
0.0085
CV imp.
0.0075 ±0.0022
Nominal p
2.10e-1
Risk diff.
-0.046

Excerpt · showing 6 of 15 expression and 6 of 15 mutation candidates.

Full sample

The full illustrative sample.

See the entire gefitinib report — every ranked indication, all 18 responder groups, both biomarker channels in full, and the disclaimer in place.

See the full sample report