DeepFCS reads flow cytometry like never before.
Every population found, gated and named — every finding traced back to its markers. No manual gating. No guesswork. In seconds.
DeepFCS is automated flow-cytometry (FCS) analysis software and a Cellular Intelligence — a multi-layered system of specialised models and agents that reads every marker on the panel, event by event, the way a cytometrist reads a population on a gate. It is decision support for haematology and immunology laboratories and research: it informs the read, it does not replace the professional.
Everything DeepFCS reads
One upload in, a fully read sample out — populations found, gated, named, and every finding traced back to the markers behind it. No manual quadrants, nothing assumed.
- Automatic gating. Populations are discovered and gated on the CD45/SSC backbone — no manual quadrants to draw.
- Population discovery & naming. Lymphocytes, monocytes, granulocytes, blasts and more are resolved and named, with counts and proportions.
- Marker-level evidence. Per-population positivity is read as weighed evidence — and every call links back to the markers that support it.
- Immunophenotyping. Full surface-marker immunophenotyping per population — antigen expression read event by event, with intensity graded and profiled against established hematology and flow-cytometry references.
- Stem & progenitor cells. CD34⁺ hematopoietic stem and progenitor populations are identified and profiled on their markers, with counts and proportions.
- Clonality & MRD. κ:λ light-chain restriction, T-cell aberrancy and minimal-residual-disease signals are read where the panel allows.
- Instrument-aware. Reads standard FCS files and adapts scaling automatically — exceptional on BD today, with more instruments on the way.
- Disease-evidence summary. Population-aware, exclusion-sensitive scoring built to EuroFlow and WHO standards — difference-from-normal, never a black box.
- Editable, expert-in-the-loop. Adjust a gate or a quadrant cursor in the built-in editors — the populations, counts and evidence re-compute, and the call stays yours.
- Reset & re-run. Undo any correction in one click — DeepFCS re-runs the classifier from scratch and rebuilds the populations, counts and evidence.
- Inspect every plot. Click any scatter to open it full-size — with its population, the markers on each axis and the exact event counts behind every gate.
How it works — from upload to result
Every step is explicit — nothing hidden, nothing assumed.
- Secure upload. Your FCS file is uploaded over an encrypted channel and stored encrypted at rest.
- Anonymization. Patient identifiers are stripped before analysis. DeepFCS never needs — or sees — who the patient is.
- Quality control. Instrument and panel detected automatically, scaling normalized, debris and doublets filtered out.
- Population discovery & auto-gating. Every population is found, gated and named on the CD45 × SSC backbone — lymph, mono, gran, blast and CD34⁺ stem / progenitor cells — with no manual quadrants.
- Immunophenotyping & marker interpretation. Every marker on the panel is read per population — positivity graded by intensity, clonality (κ:λ), aberrancy and MRD signals weighed as evidence.
- Disease evidence scoring. Population-aware, exclusion-sensitive scoring built to EuroFlow and WHO standards — difference-from-normal, never a black box.
- Review & correct. Adjust a gate or a quadrant cursor in the built-in editors — the populations, counts and evidence re-compute on the spot, and the call stays yours.
- Summary & report. A traceable disease summary and PDF — every finding linked back to its markers. In seconds.
Meet Nova — the built-in analysis assistant
Nova reads the case and explains the evidence in plain language, every claim tied back to its markers. After an analysis it starts explaining the case on its own — the diagnosis in question, the population composition, the key marker findings per tube, clonality and aberrancy — as decision support you can review and correct.
Built for the people who read cells
Wherever flow-cytometry data needs to be read quickly, consistently and transparently, DeepFCS does the heavy lifting and shows its reasoning.
- Haematology laboratories — fast, consistent first-pass reads of leukaemia/lymphoma panels, auto-gated, with marker evidence ready for review.
- Immunology & research — reproducible immunophenotyping across large cohorts, free of operator-to-operator gating drift.
- MRD monitoring — consistent identification of the populations that matter across serial samples and time points.
- Teaching & second read — a transparent, traceable second opinion, every population and finding explained back to its markers.
Security & data
- Anonymization before analysis. Patient identifiers are stripped before any processing — DeepFCS never needs, or sees, who the patient is.
- Encrypted in transit and at rest. Uploads travel over an encrypted channel and are stored encrypted.
- Hosted in Türkiye. Data is held on infrastructure located in Türkiye, in line with the Turkish KVKK and applicable data-protection law.
- Strict access controls. The analysis backend is isolated and reachable only through the authenticated application.
- You stay in control. You can request deletion of your uploaded files and derived results at any time.
- Decision support, expert-in-the-loop. DeepFCS informs, it does not decide. The clinical call remains with the qualified professional.
Frequently asked questions
- What is DeepFCS?
- DeepFCS is automated flow cytometry analysis software. It reads flow-cytometry data (FCS files), automatically finds and gates every cell population, names each one, and shows the marker-level evidence behind every finding — in seconds. It is built as decision support for laboratory and research professionals.
- Which file formats and instruments are supported?
- DeepFCS reads standard FCS files exported by flow cytometers and adapts its scaling to each instrument automatically. It is exceptional on BD today, with more instruments on the way.
- Does DeepFCS give a diagnosis or replace a pathologist?
- No. DeepFCS is a research and clinical decision-support tool. It surfaces populations, gating and marker-based evidence to assist expert interpretation, but it does not provide a medical diagnosis and must always be reviewed by a qualified professional. The final call stays yours.
- How is my data handled?
- Patient identifiers are stripped before analysis — DeepFCS never needs to know who the patient is. Files are transferred over an encrypted channel, stored encrypted, hosted in Türkiye in line with applicable data-protection law. See our Terms & Privacy.
- What can I use it for?
- Automated gating, immunophenotyping, population mapping and marker-evidence review of flow-cytometry data — across haematology and immunology research and laboratory workflows where FCS files need fast, consistent, explainable analysis.
- How long does an analysis take?
- A typical sample is read in seconds after upload — quality control, gating, population naming, marker interpretation and the evidence summary all run automatically.
About DeepFCS
DeepFCS set out to make flow cytometry readable — fast, consistent and fully transparent — so that the people who interpret cells can spend their time on judgement, not manual gating. Every result is traceable back to the markers behind it, because trust in the read matters as much as the read itself. DeepFCS is in beta and open for use while we build; each account carries a free monthly allowance, with no payment details asked.