PromptGenix
SBIR Portal

Use Cases

PromptGenix · Automated RNA-Seq + Flow Cytometry + BioAI Hypothesis Generator
What PromptGenix delivers

From raw files to reproducible results and ranked hypotheses.

PromptGenix automates core workflows used in NIH-funded labs: RNA-Seq (QC → alignment → DEG → enrichment/signatures) and flow cytometry (preprocessing → gating/subsetting → comparative analysis). On top of this, an internal LLM layer produces structured interpretation and a ranked list of testable hypotheses with suggested follow-up experiments.

PromptGenix use case montage
UMAP / t-SNE Differential expression Immune signatures Cell subset gating Gene set enrichment Hypothesis ranking
Six example panels

High-level examples (visual placeholders for SBIR web portal).

These panels illustrate the kinds of artifacts PromptGenix generates. In Phase I validation, examples will be reproduced on public datasets (GEO/SRA/FlowRepository) to avoid any data-use or IP conflicts.

Flow cytometry automated gating and subset comparison

Flow Cytometry

Automated preprocessing, subset discovery, group comparisons, and reviewer-ready plots.

  • Inputs: FCS or FlowRepository IDs
  • Outputs: subsets, markers, comparisons
Bulk RNA-Seq pipeline and differential expression

Bulk RNA-Seq

QC → alignment → counts → DEG → enrichment/signature interpretation in one reproducible run.

  • Inputs: FASTQ or GEO/SRA IDs
  • Outputs: DEG, volcano, pathways
Single-cell RNA-Seq clustering and marker discovery

Single-cell RNA-Seq

Clustering, marker discovery, cell-type annotation, and subset-level contrasts.

  • Inputs: h5ad / counts
  • Outputs: UMAP, markers, signatures
NGS automation and QC reporting

NGS Workflow Automation

Standardized ingest + QC + reporting to reduce manual glue work and re-analysis.

  • Inputs: FASTQ/metadata
  • Outputs: QC report, run logs
Gene set enrichment and immune signatures

Signatures & Enrichment

GSEA/GSVA-style outputs with immune-signature interpretation and traceability.

  • Outputs: enriched sets, rank plots
  • Use: grant/paper-ready summaries
AI hypothesis generator with citations

Hypothesis Generator

Ranked hypotheses + suggested validation experiments, grounded in observed signals and citations.

  • Outputs: hypotheses, confidence, rationale
  • Next steps: experiment suggestions
Phase I boundary: all validation examples will be generated on public datasets (GEO/SRA/FlowRepository), and the Hypothesis Generator will require evidence + citations for each ranked claim.