Data Tables
Data Tables is the spreadsheet inside Conspecta, and the place your numbers collect. Type them in, import them, or pull them from your samples and inventory, then summarize, fit, chart, and drop the result into a figure without exporting anything.
It's also where results land on their own. Every image analysis run and every set of flow statistics you send onward becomes a table here.
Creating a Table
New table opens a picker with a blank table, ten starters, any templates your team has saved, and a tile for importing a file. Choose one, decide whether the table is private to you or shared with the project, and give it a name. Nothing is saved until you've named it.
Saving a table as a template puts it in the same library as your notes and protocols, which is capped at 200 of your own and 500 shared with the team. See Templates.
The starters are structures, not data, seeded with the columns the job needs:
| Starter | Columns |
|---|---|
| Dose-response | Concentration (µM), Response A, B, C |
| Timecourse | Time (h), Condition, Measurement |
| Sample measurements | Sample, Replicate, Value, Units, Notes |
| Buffer recipe | Component, Stock, Final concentration, Volume |
| qPCR results | Sample, Target, Replicate, Cq |
| Standard curve | Standard, Known concentration, Reading 1, Reading 2 |
| Plate readings | Well, Sample, Condition, Reading |
| Cell counts | Date, Sample, Live cells, Dead cells, Viability (%) |
| Band quantification | Lane, Sample, Target signal, Loading control, Normalized |
| Experiment conditions | Condition, Variable, Level, Replicates, Notes |
You can also paste a range straight into a grid, and the table grows to fit it.
Importing
Import takes CSV, TSV, XLSX, and XLS, one table per file, named from the filename. Files are parsed off to the side so a large spreadsheet doesn't freeze the page, and rows upload in chunks so a sheet with thousands of them works.
Spreadsheet formulas survive the import. A table that arrived with formulas gets a banner explaining what happened to them, which matters because of the whitelist described below.
Cancelling an import partway deletes the half-written table rather than leaving you a mystery.
Finding a Table
The search box matches table and folder names as you type. Filter narrows what's left, and the choices combine:
| Filter | What it holds |
|---|---|
| Status | Draft, Work in Progress, Needs Review, In Review, Approved, Archived |
| Source | Linked for a table that refreshes from a project resource, Snapshot for a one time copy of a run, Typed or imported for everything else |
| Created by | One person on the project. It appears once two people have made a table here |
Folders stay visible whichever filter you pick, because the table you're after may be inside one. Switching to the list view adds a Status column, so you can read where every table stands without opening any of them.
The Grid
Everything saves as you type, and the status bar tells you where it stands. Under the grid you also get the row count, a summary of your grouping columns (something like "4 samples · 3 phyla"), and what's currently selected.
Columns come in three kinds, and the header colour tells you which:
- Raw columns are yours, typed or imported, and fully editable.
- Derived columns were produced by an Analyze action. Their cells are read-only and recompute when the values they depend on change.
- Linked columns come from elsewhere in the project and refresh from their source rather than being edited here. An amber dot means the source has moved on and a refresh is available.
Each column's menu handles renaming, sorting, filtering, formatting, width, freezing, inserting, duplicating, and deleting. Sorting and filtering are view-only: they change what you see, not what's stored, and the footer says so.
Format changes how a number is displayed, never what it is. Pick plain, a fixed number of decimals, percent, scientific, or a date style, and everything downstream, from sorting to charts to exports, still reads the underlying value. Date styles are fixed for every viewer, so a collaborator in another country reads the same column you do.
Formulas
Cells take Excel-style formulas. Start with = and the cell computes. Everything that reads a cell (sorting, filtering, charts, exports, copy and paste) reads the computed value, so a formula cell behaves exactly like the number it produces.
Selecting a formula cell brings up a formula bar, which is where you edit it. Esc reverts.
The supported functions are the ones a lab actually uses: IF, AND, OR, NOT, IFERROR, ABS, ROUND, ROUNDUP, ROUNDDOWN, SQRT, POWER, EXP, LN, LOG, LOG10, MOD, SUM, SUMIF, AVERAGE, MEDIAN, MIN, MAX, COUNT, COUNTA, STDEV, ISBLANK, CONCAT, CONCATENATE, LEFT, RIGHT, MID, LEN, TRIM, UPPER, LOWER, and TEXT.
Functions whose answer changes on its own are deliberately left
out: NOW, TODAY, RAND, INDIRECT, OFFSET and their kind.
Two people opening the same shared table would compute different numbers, and
a result you can't reproduce isn't a result. An imported formula using one of
them is kept exactly as written, showing the value it came in with, rather
than being guessed at.
Analyze
Analyze runs a named operation over the table. Pick one, fill in which columns it should use, and press Run. The result appends derived columns, a result card, or both, in a single step you can undo.
| Group | Action | What it does |
|---|---|---|
| Reshape | Summarize by group | Groups rows by a column and aggregates the rest by mean, sum, count, minimum, maximum, or standard deviation |
| Reshape | Reshape | Moves between wide and long layouts |
| Transform | Normalize to a reference | Divides a column by another column or a fixed value |
| Transform | Percent change vs a baseline | Expresses values as a percentage change from a baseline group's mean |
| Compare | Correlate | Pearson or Spearman correlation between two columns, with r, r², and n |
| Compare | Compare groups | A t-test or ANOVA of one column across groups, with per-group means and significance |
Results that aren't columns appear as cards in the Results panel and are saved with the table.
Assays
Three guided fits sit in the same menu, under Assays:
- qPCR ΔΔCt normalizes target Ct values to housekeeping genes and a calibrator group, producing ΔCt, ΔΔCt, and fold-change columns
- ELISA standard curve (4PL) fits a standard curve from your knowns and interpolates the unknowns into a concentration column
- Dose-response IC50/EC50 (4PL) fits a dose-response curve and reports the IC50 or EC50
A table can also be created as a qPCR assay, in which case it opens on a short setup that maps your columns and previews the ΔΔCt calculation before committing. The preview is the real computation, so what you review is exactly what lands.
Assay tables draw their fitted curve over a scatter chart of the same data, so the fit and the points are one picture.
Charting
Chart opens a drawer under the grid. It has four controls: chart type, the x axis, which columns are values, and what to group by. Leave any of them empty for a sensible default.
This is a preview, not the chart editor. When it looks right, Add to Figure promotes it into an existing figure or a new one, and the full styling controls are there. Export takes the preview itself as PNG, SVG, or TIFF at a printed width in millimetres and up to 600 DPI, or copies it to the clipboard. The SVG is real vector: bars, points and axis ticks each select on their own and the legend is live text.
A rail beside the chart lists every figure already built from this table, each one a link.
On a phone the chart takes the whole screen rather than sharing it with the grid. Held upright, the four controls fold into a single line naming what's charted, which you tap to open; turned sideways they sit beside the chart as usual.
Embedding Live Data
Embed data pulls a project resource in as columns that refresh from their source. It adds a row per item and leaves your own columns and rows alone.
Two kinds of source:
Live resources bring in a whole workspace at once, one row each: Samples, Inventory, Orders, Tasks, Locations, Genetics, and Image analysis. Sample status and health, stock quantities, order totals, task deadlines: all of it refreshes from the workspace that owns it.
Assays and plates bring in one specific thing you pick: a Plate, with a row per well, a Sample set, with a row per member, or one or more Analysis runs, with their per-image results. Pick several runs and a Group column names each row's source, so treated and control land side by side ready to chart.
A linked table fills once and is then yours. Add your own columns beside the pulled ones, annotate rows, and refresh when you want the latest. If an upstream record is deleted, its row stays with the values you last pulled and a flag saying the source is gone, rather than vanishing and taking your notes with it.
Finding Things
Search table filters rows to those matching in any column, and a caret in the search field opens find and replace. Group by folds rows into groups by a column's value. Freeze up to here on a column keeps the identifying columns on screen while you scroll right.
Cmd/Ctrl + Z and Cmd/Ctrl + Shift + Z undo and redo grid edits.
Sharing and Collaboration
A table is private to you or shared with the project, chosen when you create it and changeable later.
A shared table in a team project supports live co-editing: two people can work in the same grid, with presence avatars in the toolbar showing who's in it. Very large tables (past about 50,000 cells) turn live editing off and stay fully editable by one person at a time, since keeping a sheet that size in sync would cost more than it's worth.
Project members with view-only access can read and export any table but can't change it.
Tables carry a status from Draft through to Approved or Archived, and an Activity panel with comments and sign-offs.
Exporting
Export as Excel is lossless: formulas, number types, and formatting all survive. Export as CSV writes values only. Both export the table as it currently stands, including edits you just made and values you just refreshed.