An anecdote, some numbers, and a pressing question
I remember the autumn of 2017 in my small Athens lab, when a 96-well spatial transcriptomics run lost 40% of its usable regions — a bruise that taught me more than textbooks ever did; and yet the landscape of spatial technology companies kept promising seamless pipelines. Spatial omics service was the term every vendor used in slides and brochures, but practice and promise often diverged. Scenario: a pilot study with precious biopsy material. Data: 12 samples, one month, nearly half the data degraded. Can we build services that respect the sample as much as the science?
What went wrong?
I have run assays where multiplexed imaging signals faded between runs, and I have overseen single-cell captures that collapsed under poor tissue handling. I speak from direct, repeated encounters: in 2019 I shipped slides overnight from Thessaloniki to a partner lab and a humidity swing rendered two slides unusable — a concrete loss, a grant deadline missed. The traditional fixes — tighter SOPs, more QC checkpoints — are necessary but insufficient. Vendors often treat spatial transcriptomics like merely another sequencing add-on; they under-resource sample logistics and ignore the human work at the bench (frankly, that always annoyed me). The deeper flaw is systemic: workflows built around instrument throughput, not sample fragility, breed hidden pain points for end users and wholesale buyers alike.
Root causes and user pains
I can list root causes from experience: poorly standardized tissue embedding, pipelines that assume ideal RNA integrity, and vendor tools that prioritize throughput over reproducibility. These translate into the client-facing pains I have seen — unpredictable turnaround, opaque QC criteria, and extra costs for repeat runs. We often found that multiplexed imaging protocols were applied as if all tissues were identical; they are not. Single-cell resolution demands bespoke handling. My team once adjusted a fixative protocol for liver biopsies and recovered signal in 8 of 10 samples — a small protocol change, large impact. That detail is the kind of thing sellers rarely highlight, but I always share it; you know, plain truth.
Comparative future — a technical shift in perspective
Now, moving to a forward-looking, technical posture: spatial technology must evolve from instrument-centric to context-aware platforms. I look to vendors who integrate sample tracking, real-time RNA integrity metrics, and adaptable wet-lab workflows. When I compare three providers I evaluated in 2022, the winner was not the one with the fanciest imager but the one that provided per-sample metadata, automated in situ hybridization QC flags, and transparent failure modes. This is comparative thinking — not promotional — and it changes procurement criteria for wholesale buyers and lab managers. (Short aside: automation matters, but not at the cost of local expertise.)
What’s Next
Adoption will hinge on interoperability: instruments must output standardized metadata so downstream analysis — be it spatial transcriptomics maps or multiplex protein panels — is reproducible across sites. I advise teams to insist on traceable chains of custody for samples, on vendor training tied to performance metrics, and on pilot agreements that quantify sample recovery. I once negotiated a pilot in 2020 that capped liability to a sample loss threshold; it forced the provider to tighten procedures — measurable change. Interruptions happen — we adapt. Then we measure again.
Three practical evaluation metrics I use
I conclude with three concrete metrics I require before I commit to a long-term provider relationship: 1) per-sample integrity recovery rate (report historical % of samples meeting QC), 2) metadata completeness (time-stamped handling, reagent lot IDs, and in situ QC flags), and 3) reproducibility score across technical replicates (expressed as variance in gene detection per region). These metrics let me compare offers honestly and protect scarce samples. For practical procurement and collaboration, those numbers matter more than glossy brochures. In closing — a small human note — I still believe that careful methods win more battles than flash. For partners and procurement folks, I often point them to one pragmatic place: stomics.