OV Platform Data Portal

Internal research portal for Sator's precision oncolytic virotherapy platform. Eight years of systematic screening, interactive analytics, and generated reports.

6,317
Data Points
38
Viruses
96
Cell Lines
9
Cancer Types
16
Receptor Genes
1,064
Predictions

Screening Platform

Interactive tools for the 96-line × 38-virus screening panel. 6,317 data points across 9 cancer types, 8 years (2018–2026).

2018–2026

OV Kill & Receptor Heatmap

96 cell lines × 38 viruses. Toggle basal/+IFNβ, predictions, IFN sensitivity. Receptor expression with 6-tier data sourcing. Chumakov 2018 IFN integrity panel (28 genes). Colorblind-safe mode.

A-01HTMLCONFIDENTIALSCI, INV-D📊 6,317 data points🧬 16 + 28 genes
2018–2026

Dilution Profile Explorer

Dose-response curves for every virus × cell line combination. Singles across all 103 lines; 11 multi-virus combos run on 10 mesothelioma + control lines only. 7 dilution points (10⁻² to 10⁻⁸), basal and +IFNβ. Compare profiles, overlay curves, filter by cancer type.

A-05HTMLCONFIDENTIALSCI, INV-D📈 7,793 profiles🧪 11 combos · 220 rows
2019–2021

IFN-β Dose-Response Profiling

Standalone experiment. 46 cell lines ranked by IFN-β fold-change. Dose-response sparklines. 4-tier sensitivity zones. 4 corrected from raw plate data.

A-04HTMLCONFIDENTIALSCI, PUB🧫 46 lines📈 5 cancer types
Oct 2021

Combination OV Viewer

4-virus pairwise combinations on mesothelioma. Kill heatmap, dose-response curves, Bliss synergy analysis, PBMC safety.

A-02HTMLCONFIDENTIALSCI🧪 6 combos📉 Dose-response
2018–2026

Scorecard / Clinician / Poster

Three simplified views of the primary screen: traffic-light scorecard for executives, ranked virus selector for clinicians, condensed poster heatmap.

A-03HTMLCONFIDENTIALSCI, INV-D🎯 3 audience modes
2019–2021

IFN-β Dose-Response Profiling

Dedicated experiment. 46 cell lines, 5 cancer types. Dose-response fold-change ranking. 4 values corrected from raw plate data. Zone classification.

A-06PDFCONFIDENTIALSCI, PUB🧫 46 lines📊 4 corrected
2019–2021

OV Screen IFN-β Impact Analysis

Per-cancer-type IFN analysis, CDx implications, monotherapy vs. combination strategy, indication prioritization.

A-07PDFCONFIDENTIALSCI, INV-D🧬 CDx thesis
Oct 2021

Meso Combination Screen

4 lead viruses × all pairwise combos. Additive basal, dramatic synergy under IFN (+69pp). CoxA7 PBMC-safe.

A-08PDFCONFIDENTIALSCI🧪 6 pairwise

Expression & Pathway Biology

First-party RNA-seq, receptor correlations, IFN integrity mapping, protease processing, and molecular characterization.

2024–2026

Receptor-of-Entry Heatmap

16 genes (15 receptors + CXCL10) across 96 cell lines. TPM, binary, z-score views. Chumakov 2018 IFN integrity toggle (28 genes). Colorblind-safe Cividis mode. 6-tier data sourcing.

B-01HTMLCONFIDENTIALSCI, PUB🧬 16 + 28 genes × 96 lines📊 4 view modes
2018–2019

RNA-seq Expression Panel

First-party RNA-seq: Mesothelioma (7 lines + 2 LP-9) and Ovarian (11 lines + OEpi). FPKM expression, log₂FC, DESeq2. Receptor genes, IFN response, and Chumakov 2018 IFN Integrity (28 genes).

B-03HTMLCONFIDENTIALSCI, PUB🧫 21 lines📈 3 panels × 2 tabs
2018–2019

RNA-seq Analysis Report

First-party RNA-seq: mesothelioma (7 + LP-9, Set B) and ovarian (11 + OEpi, Sets E/F). F11R/ITGB3 CCLE discrepancies, ISG induction, Chumakov 28-gene panel, QC corrections (label swaps, EKAZ). Cross-panel findings.

B-04PDFCONFIDENTIALSCI, PUB🧬 21 lines📈 3 panels
2026

IFN Pathway Integrity (Chumakov 2018)

28-gene IFN integrity panel across 64 lines (20 RNA-seq + 44 CCLE). cGAS silenced 23% pan-cancer. IRF9 80% depleted. Three-tier CDx. DES calibration.

B-06PDFINTERNALSCI🧬 28 genes🧫 64 lines
2019–2020

IP10/CXCL10 Investigation

ELISA + WB. r = −0.028 — not predictive of OV kill. Reinterpreted as immune contexture marker for OV + anti-PD-1 combination.

F-05PDFCONFIDENTIALSCI🔬 ELISA + WB
2026

Sendai Protease Analysis

5-gene protease panel across 61 lines. STR0028 (Sendai Moscow) protease-independent — kills regardless. STR0030 (Sendai Cantell) requires proteases and fails in protease-negative lines.

B-07PDFCONFIDENTIALSCI🧬 5 genes × 61 lines
2026

Expression Source Divergence

Why Sator RNA-seq and CCLE produce different prognostic signals. Culture condition effects on metabolic genes, genetic drift (Ben-David 2018), receptor stability vs metabolic dynamics. 57% R²adj loss quantified.

F-13PDFCONFIDENTIALSCI, INV-D📄 4 pages

Prognostic & Resistance

Pathway-centric resistance analysis, gene pair CDx candidates, cancer-type-specific drivers, literature validation, and treatment strategies for resistant tumors.

2026

Prognostic Biomarker Explorer

Genome-wide resistance analysis: 189 pathway genes × 69 cell lines × 38 viruses. 17 viral lifecycle pathways, gene pair predictions (R²adj=0.68), cancer-type-specific drivers, per-cell-line resistance signatures. Literature IFN combos tested. Cividis colorblind mode.

B-07HTMLCONFIDENTIALSCI, INV-D🧬 189 genes × 69 lines📊 6 interactive tabs
2026

Prognostic Explorer — CCLE Only

Same 189-gene × 17-pathway analysis using only CCLE/DepMap expression data. Independent validation of prognostic findings without Sator RNA-seq. Zero pathways reach significance in CCLE-only (vs 5 in mixed-source) — the Sator RNA-seq provides unique predictive power.

B-07cHTMLCONFIDENTIALSCI🧬 CCLE-only expression📊 6 tabs
2026

Literature Validation Report

Peer-reviewed validation of all Prognostic Explorer findings against 30 published studies. 17 pathways scored, 7 cancer-specific markers checked, metabolic reprogramming and gene pair models assessed. 4 Chart.js interactive charts, clickable PMC PDF links, graduate-level plain-language explainers.

B-08HTMLCONFIDENTIALSCI, INV-D, PUB📖 30 citations📊 4 charts · 6 tabs
2026

CCLE-Only Literature Validation

Side-by-side comparison of CCLE-only (60 lines) vs mixed-source (69 lines) prognostic explorers. 57% R²adj drop without Sator RNA-seq. Autophagy signal collapses (p=0.003→0.868). MCL1 melanoma confirmed across both. 5 tabs: pathway comparison, cancer-type markers, cross-virus stability.

B-08cHTMLCONFIDENTIALSCI, INV-D📊 5 tabs🔬 60 vs 69 lines
2026

Expression Source Divergence

Quantitative comparison of 177 genes in the prognostic panel between CCLE (n=60) and Sator RNA-seq (n=21). 91 genes (51%) change correlation direction between platforms. 8 genes show major disagreements (|Δρ| > 0.5). Identifies which findings are platform-stable vs condition-dependent. Interactive sortable table with 3 tabs: divergent genes, methodology, implications.

B-08HTMLCONFIDENTIALSCI, PUB🧬 177 genes⚠️ 91 flips (51%)
2026

Prognostic Analysis — Technical

Full statistical methods: Spearman correlations, FDR correction, pairwise regression, Mann-Whitney DE, pathway activity scores. All tables with p-values and effect sizes. Overfitting assessment. IFN-β response analysis.

F-11bPDFCONFIDENTIALSCI📄 4 pages
2026

Prognostic Analysis — Plain Language

Graduate-level narrative: why autophagy/trafficking drive OV resistance (not IFN). Gene pairs as CDx candidates. Cancer-type-specific therapeutic strategies. Limitations and next steps.

F-11aPDFCONFIDENTIALSCI, INV-D📖 Guide
2023

Treatment Strategy

Virus selection framework per cancer type. Cocktail design rationale. Clinical pathway mapping.

F-08PDFCONFIDENTIALSCI, INV-D📋 Strategy
2026

OV-Resistant Lines

Why 13 tumors resist most viruses and how to overcome each one. Part 1: Molecular diagnosis via Chumakov 28-gene panel. Part 2: Individual treatment plans — sensor gaps, JAKi combos, sequential dosing.

F-09PDFCONFIDENTIALSCI, INV-D🧫 13 lines🛡️ Resistant
2026

Prognostic Explorer Comparison

Mixed-source (n=20) vs CCLE-only (n=60). Pathway p-value divergence, FCGRT+SOCS1 cross-validated pair, metabolic signal collapse. Side-by-side tables.

B-07dPDFCONFIDENTIALSCI📄 3 pages

Predictions & Validation

KNN receptor-expression model. 1,064 locked predictions (532 basal + 532 IFNβ). LOOCV-validated, SHA256-timestamped.

Preclinical & Manufacturing

In vivo evidence, production optimization, and manufacturing feasibility.

Investor & Business

Market opportunity, go-to-market planning, and investor-ready materials.

Plain-Language Guides

Non-specialist explanations of every major component. Written for investors and collaborators.

2026

Prognostic Explorer Guide

Tab-by-tab walkthrough of the interactive visualizer. Controls (dark mode, colorblind, virus names), sidebar filters, heatmap column drag, gene combinations, cancer type resistance, cell line profiles. How to read all statistics.

F-12PDFCONFIDENTIALSCI, INV-D📖 Guide
2026

OV Dashboard Guide

How to navigate the kill heatmap, receptor expression, and compare tabs. Toggle buttons, sorting, IFN modes, predictions overlay, colorblind mode, and practical tips.

G-12PDFCONFIDENTIALINV-D, INV-L📖 Guide
2026

Dilution Explorer Guide

How to read dose-response curves, compare profiles, understand curve shapes, assess potency, and interpret combination synergy data.

G-13PDFCONFIDENTIALINV-D, INV-L📖 Guide
2026

CDx Framework Guide

The four-layer companion diagnostic: receptors, IFN integrity, phenotypic assay, proteases. How we match viruses to tumors.

B-02gPDFCONFIDENTIALINV-D, INV-L📖 Guide
2026

KNN Model Walkthrough

Step-by-step prediction algorithm. 16-gene receptor panel, z-score normalization, K=3, inverse-distance weighting, LOOCV validation.

C-05PDFCONFIDENTIALINV-D, INV-L📖 Guide
2026

IFN Screening Guide

Why we test every virus twice. Sensitivity zones, cancer-type patterns, the immune primer paradox, virus selection implications.

A-07gPDFCONFIDENTIALINV-D, INV-L📖 Guide
2026

IFN Integrity Guide

How cancer cells lose antiviral defenses. The relay race: detection, signaling, response. cGAS silencing, MDA5 variation, IRF9 bottleneck.

B-06gPDFCONFIDENTIALINV-D, INV-L📖 Guide

Methods & Reference

Protocols, methodology documentation, and reference materials.

Document Classification

Every asset in this portal carries a classification badge. This legend defines the system.

Distribution Tiers

INTERNALFor internal use only. Build tools, audit scripts, raw data, working notes. Never leaves the team.
CONFIDENTIALShared under NDA or at founder discretion. Investor materials, analytical reports, preclinical evidence. Always know who has a copy.
DISTRIBUTABLECleared for external distribution. Published manuscripts, preprints, supplementary data. No restrictions.

Audience Tags

SCIScientific / Technical — PhD-level reviewers, journal editors, collaborators, scientific advisors.
INV-DInvestor (Deep) — Life science VCs, pharma BD, scientific due diligence teams.
INV-LInvestor (Light) — Angels, family offices, generalist VCs wanting thesis-level narrative.
INTInternal / Operational — Team members, build tools, CI/CD.
PUBPublication / Peer Review — Journal submission, preprint, supplementary materials.
VETVeterinary / Translational — Vet oncologists, canine case study reviewers.

Asset Categories

A-##Screening Data & Analysis — Core OV kill data, IFN sensitivity, screen results.
B-##Receptor Biology & Expression — Gene expression, receptor mapping, RNA-seq.
C-##Predictive Model — KNN imputation, validation, predicted heatmaps.
D-##Preclinical Evidence — In vivo studies, patient data, xenograft results.
E-##Investor Materials — Portal, pitch deck, screeners, audience-segmented deliverables.
F-##Publication & Records — Manuscript, analytical reports, production records.
G-##Infrastructure & Standards — Templates, design tokens, audit tools, portal itself.

Status

LIVECurrent, deployed, and up to date.
STALEExists but needs update — data, naming, or structural changes pending.
PENDINGPlanned but not yet built.
BLOCKEDWaiting on external dependency.
37 registered assets· 14 INTERNAL· 21 CONFIDENTIAL· 2 DISTRIBUTABLE· 7 categories (A–G)· 6 audience tags