Validation
What GENESI-X1 is, and what it is not
Nov 21, 2026 · 3 min read · GeneSI Lab
Everything we report is computational. This note sets out the scope of v1.0, the limitations we state openly, and what must come next.
Clear scientific reporting requires stating what a system cannot do just as plainly as what it achieves in simulation. This note defines the exact scope of GENESI-X1 v1.0, lists its five stated limitations and outlines the experimental and engineering steps required next.
What it is
GENESI-X1 v1.0 (Genomic Engine for Nucleic-acid Evaluation, Specificity, and Intelligent Assay Design, Generation X, version 1.0) is an in-silico research prototype built by GeneSI Lab. It pairs seven modules (IngestCore, VarShield, CandGen, BlastSpec with ViennaRNA biophysics, ThermoSpec-LM, MultiPlexOpt and ExplainX) with ReportForge outputs and a Streamlit prototype interface.
The technical report documenting v1.0 is a pre-peer-review computational report. It demonstrates how variant-aware candidate generation, BLAST specificity screening, learned ranking and simulated annealing multiplex optimization behave across 24 simulated benchmark loci.
What it is not
GENESI-X1 v1.0 is not a medical device and has not been evaluated for diagnostic or regulatory use. No wet-lab experiments, no patient tests and no clinical validation have been performed on any primer or probe sequence proposed by the software.
Every metric reported for v1.0, from Reference Quality Score (RQS) improvements to diploid dropout estimates and 8-plex heterodimer counts, is strictly computational (in silico). Proposed oligonucleotide sequences must be synthesized and benchmarked in a laboratory before anyone can rely on them physically.
Five limitations we state openly
We document five explicit methodological, statistical and licensing boundaries in the GENESI-X1 v1.0 report:
- Author-defined rubric and input overlap: 24 of the 78 ranker features enter RQS directly. True assay value must be measured against wet-lab Cq, efficiency and limit of detection (LoD).
- Limited 8-gene statistical power: a pooled 15-test correction is not significant (p = 0.1172), some exposure gaps lose significance, and B3 leads on specificity alone (0.7969 versus 0.7510).
- Transcriptome-focused BLAST index: BlastSpec screens 186,716 RefSeq RNAs plus target loci and 11 paralogs, leaving deep intronic or intergenic off-targets unscreened.
- Non-commercial model license: the 78-feature configuration includes frozen Nucleotide Transformer NT-v2-50M embeddings under CC BY-NC-SA 4.0, restricting commercial use unless those embeddings are omitted.
- Zero wet-lab validation: no primers or probes from v1.0 have yet been synthesized, amplified or sequenced at the bench.
What must come next
First, we plan to execute a MIQE-aligned wet-lab validation synthesizing HPLC-purified primers and dual-labeled hydrolysis probes for our proposed 8-plex panel. Bench runs will measure real efficiency, linearity, paralog specificity, variant stress resistance and LoD.
Second, in planned GENESI-X1.5 and GENESI-X2 releases, we plan to extend specificity screening to the full GRCh38 primary assembly. We also plan to add ancestry-stratified filtering, adopt exact multiplex solvers and calibrate rankings against empirical qPCR curves.
These results are computational (in silico). GENESI-X1 v1.0 is a research prototype, not a medical device, and no wet-lab or clinical validation has been performed. For research use only.
Related
Questions about this work? Email contact@genesilabs.xyz