SuperTruth xMD Anderson Cancer Center

VIOLET

Visual Intelligence Layer for Oncology Trends & Evidence Triangulation

3D semantic maps of 351 CLL patients — clustered by mutation profiles, methylation burden, and treatment outcomes. Search across the entire Wierda dataset by gene, subtype, or clinical marker.

351 CLL patients mapped
10 gene mutation profiles
48 CpG methylation sites
5 leukemia subtypes clustered

Mutation Heatmap — 10 Patients x 10 Genes

KRAS
RUNX1
NPM1
FLT3
TET2
TP53
DNMT3A
IDH1
IDH2
ASXL1
WRD-001
CLL
WRD-002
AML
WRD-003
CLL
WRD-004
ALL
WRD-005
CML
WRD-006
CLL
WRD-007
AML
WRD-008
CLL
WRD-009
MDS
WRD-010
CLL
Mutated
Wild type

Wierda CLL Dataset — Data Fields

Gene DNA mutations

KRAS, RUNX1, NPM1, FLT3, TET2, TP53, DNMT3A, IDH1, IDH2, ASXL1 — binary mutation status per patient. Co-mutation matrices generated automatically.

Gene RNA expression

CSF3R, ABL1, CALR, IDH2, JAK2, MPL — RNA-level gene expression markers. Differential expression across subtypes visualized.

48 CpG methylation sites

Beta values (0-1) for each CpG site from Illumina 450K and EPIC arrays. Heatmaps, clustering, and burden scores generated per patient.

Leukemia score

Composite leukemia confirmation score derived from blast %, lab values, and genetic markers. Used for cohort segmentation and treatment response correlation.

Subtype clustering

5 leukemia subtypes (AML, ALL, CLL, CML, MDS-MPN) as primary clusters. Sub-clusters from mutation profiles and methylation patterns.

Treatment outcomes

Protocol assignments, response categories (CR, PR, PD), survival status, and relapse events — correlated with genomic markers.

Data Pipeline — Ingest to Visualization

01

Ingest

Wierda CLL dataset — 351 patients, 96 fields

02

Profile

10 gene mutations + 48 CpG sites per patient

03

Cluster

3D embedding by subtype, mutation, methylation

04

Correlate

Treatment outcomes × genomic markers

05

Visualize

Interactive 3D maps + searchable cohort

Platform Capabilities

3D semantic cohort maps

351 CLL patients plotted in 3D space — clustered by subtype, gene mutations, methylation burden, and treatment outcomes.

Gene mutation network

DNA mutation profiles visualized as interconnected networks. Co-occurrence patterns surface actionable mutation combinations.

Methylation burden analysis

48 CpG sites analyzed per patient. Epigenetic patterns correlated with treatment response and survival outcomes.

Evidence triangulation

Search by mutation, blast %, subtype — matching patients light up. Cross-reference with CLL research literature.

Treatment trend tracking

Protocol assignments, response rates, and outcome distributions visualized across the entire cohort.

Literature integration

CLL research papers indexed alongside patient data. When patterns emerge, VIOLET surfaces supporting published evidence.

See the live platform

Explore VIOLET's interactive 3D semantic maps — search across 351 CLL patients by mutation, subtype, methylation burden, and treatment outcomes.

Launch VIOLET