Tumor models, immune phenotyping, and tissue imaging in one project
A preclinical cancer study produces caliper measurements, FCS files, stained sections, and a survival curve, usually in four different tools. Conspecta keeps all of it attached to the animal and the treatment arm it came from.
The preclinical cancer workflow, in one browser tab
One project. Every caliper series, FCS file, stained section, and figure stays linked to the animal, treatment arm, timepoint, and model behind it.
Growth
Record volumes and endpoints in a data table attached to the same animal that its flow files and tissue images hang off, so a growth curve and an endpoint phenotype describe one mouse.
How most labs do it today. The growth curve lives in a file that has no idea which animal produced which endpoint sample. Linking a fast-growing tumor back to its flow data is a manual lookup.
Phenotyping
Compensation and multicolor gating in the browser, saved as a template the whole lab gates from, with every FCS file linked to the tumor it was dissociated from.
How most labs do it today. Gating happens in a workspace file on one laptop, under a per-seat licence. The arm and timepoint of each sample live somewhere else entirely.
Clustering
SOM and graph clustering (FlowSOM-style and PhenoGraph-style), K-means, and DBSCAN, plus UMAP, t-SNE, and PCA, running in the browser with nothing to install.
How most labs do it today. A plugin install per machine, or a bioinformatician in the loop, for a clustering result you wanted before the next cohort goes in.
Imaging
Open an image and a segmentation model counts every cell, with no threshold to tune. Marker intensity, area, and colocalization (Pearson's, Manders', Costes) on LIF, TIFF, and PNG fields, with one detection setup reused across the whole cohort.
How most labs do it today. Infiltrating cells are counted by hand, and the threshold used last season isn't written down beside the counts, so a repeat starts by re-deriving it.
Plots
Charts are built from the gated populations and image measurements themselves, so re-running the analysis moves the numbers downstream instead of orphaning them.
How most labs do it today. Every re-gate or re-threshold means a fresh export and a fresh paste, and the chart that results has no route back to the analysis behind it.
Cohorts
Each sample carries its treatment arm, timepoint, and model, and every file measured from it inherits that context into every plot and figure.
How most labs do it today. Which sample was the anti-PD-1 arm, day 21, cage 3? The answer is a filename convention that holds until the study grows a second cohort.
Writing
Notes, flow analysis, image analysis, and multi-panel figures sit in one project, and the figure is built from the analysis rather than pasted beside it.
How most labs do it today. The figure is a screenshot of a plot that has been re-analyzed twice since, and the notebook describes the gate in prose.
See it on your own cohort
Bring one tumor panel or one stained section and see the analysis come back attached to the data it ran on.