Titan-1™ technology
The long read behind every GutID report.
Titan-1 combines an in-house lysis and extraction step with a patented amplicon and our own classification software.
Region proportions are schematic.
Published in peer-reviewed journals
- Microbial Genomics
- mBio
- Scientific Reports
- PLOS ONE
- Frontiers in Microbiology
- ISME Communications
- Communications Biology
- International Journal of Molecular Sciences
The sequence
What happens to your sample in the lab.
Titan-1 is targeted sequencing. It reads one informative region on purpose, instead of sequencing everything present and sorting it out afterward.
It starts with the sample
A stool sample carries hundreds of bacterial species. The first job is to capture as many of them as possible.
Gentle, thorough lysis
A proprietary lysis and extraction step, made in-house, opens tough-walled organisms, many Gram-positive species among them, while keeping DNA intact across the more fragile ones.
One patented amplicon
A patented step generates one fragment of roughly 2.5 kilobases spanning the full 16S gene, the ITS and the start of 23S: conserved stretches that place an organism, hypervariable ones that separate close relatives.
Consensus from repeated passes
The whole fragment is carried in one continuous long read, and each molecule is read repeatedly so the consensus built from the passes is more accurate than any single pass.
Not limited to a reference database
Our own classifier identifies organisms from the whole long read instead of matching short markers to a fixed database.
Read as an ecosystem
Together, these steps turn strain-level data into species-level identification with relative abundance, so the report can show how much of the community each organism holds.
Diagrams are schematic, not data.
Side by side
How Titan-1 compares.
Every sequencing approach trades how much of the community it sees against how precisely it can identify what it finds.
| qPCR | 16S-only | Shotgun metagenomics | Titan-1 | |
|---|---|---|---|---|
| What is read | A fixed panel of chosen targets | One short variable stretch of the 16S gene | Everything present, in short fragments | The full 16S gene, the ITS and the start of 23S, in one ~2,500-base read |
| Targeted or untargeted | Targeted, but only at what you already suspect | Targeted, but short | Untargeted; depth is spread across every gene present | Targeted, and complete across the region |
| Typical resolution | Only the organisms on the panel | Genus, sometimes species | Genus to species; strain-level calls need deep sequencing | Species, from strain-level data |
| Unfamiliar organisms | Not detected; absent from the panel | Often unresolved or mis-assigned | Depends on the reference database | Separated and quantified even when no species name exists |
| Relative abundance | Not comparable across targets | Shifted by primer choice on a short region | Affected by genome size and fragmentation | One continuous target, so less fragmentation bias |
| Analysis burden | Minimal, but the answer is narrow | Small | Very large | Focused on one bacterial target |
Rows describe the general characteristics of each method class, not any specific commercial product. Comparison informed in part by peer-reviewed work: Gehrig et al., Microbial Genomics, 2022.
What it produces
Reading a Target Plot, ring by ring.
Each ring of the Target Plot is a level of classification: all bacteria at the center, then phylum outward to species at the edge. Every band is sized by how much of the sample that organism accounts for, so the whole figure is drawn to scale.
See the method on your own sample.
Collected at home in about five minutes, and read with the method described on this page.
