Make each BioProfile result traceable to its culture collection
Seal can bring culture chemistry and cell measurements into the same experiment record as vessel history and sample preparation. This guide works through a completed-file handoff, two overlapping exports and a cell-density mapping. The receiving design is illustrative and must be implemented and tested with your installed FLEX2 software.
Choose the source route
Start with a selected historical export
The FLEX2 manual describes exporting selected Historical Results to the Bridge or USB. Sample results, QC results and images have separate exports. Nova: BioProfile FLEX2 instructions for use, §§2.2.6, 2.3.3 and 3.6.4.
In Seal: Take a small completed sample-result file first. Retain the original file and its source selection. Arrange delivery to the IoT computer’s approved watched folder and verify one import before scheduling transfers.
Give the monthly file a completed handoff
Auto Export appends new results to a monthly CSV on the Bridge. Nova: BioProfile FLEX2 instructions for use, §§2.2.6, 2.3.3 and 3.6.4.
In Seal: Use a site-owned transfer job to produce a consistent, closed snapshot before collection. Define how it detects an incomplete copy and reconciles overlapping rows. A file watcher does not itself implement row-level ingestion or replay recovery.
Retain reanalysis relationships
An edited or reanalysed result receives a new historical entry; Sample Time retains the original entry’s timestamp. The pre-dilution multiplier affects CDV results. Nova: BioProfile FLEX2 instructions for use, §§2.2.6, 2.3.3 and 3.6.4.
In Seal: Keep the source evaluation relationship and independently recorded culture collection time. Inspect the actual exported fields with the instrument owner; a same-named sample or a changed timestamp alone cannot establish a new collection.
Reconcile two exports before scheduling culture-result collection
Use two test analyses with known sample, vessel and preparation records. Obtain the original exports, the matching analyzer reports and the collection times. The first run should prove the quantity mapping and replay behavior with a small population that can be checked manually.
Download the first-run procedure and receiving checks
Choose the source population and supporting records
Select the two intended analyses in Historical Results, checking the displayed date range. Export the chosen sample results and retain the matching reports. Collect any required QC evidence and cell images separately. Record the installed software and the exact export headers before mapping.
Check: The packet accounts for both analyses and identifies which sample each represents. Missing QC or image evidence remains visible; a sample-result CSV alone does not prove that those separate artifacts were collected.
Nova: BioProfile FLEX2 instructions for use, §§2.2.6, 2.3.3 and 3.6.4
Make collection time and source identity explicit
Match each export entry to the analyzer display and the independently recorded culture collection. Record how the source identifies repeat measurements and later evaluations. Use A-101 and A-102 only as local worksheet labels until the real source identity contract has been established.
Check: The proposed key distinguishes two legitimate measurements with the same sample name. A re-export resolves to the existing entry. If the export cannot unambiguously establish that relationship, retain an unresolved mapping rather than inventing identity from row order.
Check a density, a percentage and a chemistry result
Work through the selected analysis below with its report and preparation record. Read the density scale from the exported unit, establish which corrections the reported result already includes, and map glucose independently. Retain any missing or qualified result with its status.
Check: For the fictional case: density is 4,800,000 cells/mL, viability is 96%, and glucose is 3.2 g/L. Lactate has no numeric value. Save the original representation and transformation evidence with each receiving field; one sample-level factor is insufficient to define every quantity.
Deliver closed files with deliberate overlap
Make one test handoff containing A-101 and a second containing A-101 plus A-102. The transfer owner must define how a consistent export or snapshot becomes a closed file before it enters the watched folder. Include a test where transfer stops halfway; temporary files must not be presented as complete handoffs.
Check: After both completed handoffs, the result population contains two analyses, with the repeated delivery attributable. Compare expected source entries, parsed entries, unresolved entries and destination identities. An incomplete transfer should be recoverable without claiming that the missing portion was imported.
Check recovery and a later evaluation
Interrupt the receiving process after A-101 is retained, then deliver the second handoff again. Separately include a supported later evaluation of A-101’s sample and retain its relationship to the original analysis. Compare the destination history and culture timeline after both tests.
Check: Recovery completes the intended population without creating another A-101. The later evaluation remains identifiable under its original collection; it does not silently replace earlier evidence or move the culture’s collection time.
Inspect quantities independently
These are fictional mapping decisions. Use the actual export units and correction evidence; the instrument’s preparation settings do not define every destination field by themselves.
Scroll across to see every column →
| Field | Source evidence | Receiving result |
|---|---|---|
| Viable cell density | 4.8 × 10^6 cells/mL; correction included | 4,800,000 cells/mL; no second preparation factor |
| Viability | 96% | 96% in a percentage field, not 96 cells/mL |
| Glucose | 3.2 g/L | 3.2 g/L under its confirmed chemistry basis |
| Lactate | Unavailable | No numeric value; preserve status |
| Culture collection | S-301 collected 09:00 UTC | Plot at the collection event; retain analysis at 09:18 separately |
Expected receiving ledger for the two handoffs
Illustrative reconciliation CSV, not Nova’s export format. The actual connector must supply the identity matching and recovery behavior represented here.
handoff,source_analysis,receiving_action,expected_analysis_count
handoff-01.csv,A-101,create,1
handoff-02.csv,A-101,link repeated delivery,1
handoff-02.csv,A-102,create,2The ledger has three source entries and two scientific analyses. Keep delivery history as evidence without counting each arrival as another culture measurement. If A-101 changes under the same claimed identity, resolve that conflict before treating the entry as a harmless repeat.
Two snapshots contain three source entries, but only two analyses
Fictional normalized records, not Nova CSV headers or built-in IDs. A-101 and A-102 represent source identities established from the installed export contract. The preparation check explicitly confirms that reported cell density already includes the applicable correction.
Scroll across to see every column →
| Source information | Example identity or value | Why it stays distinct |
|---|---|---|
| Culture position | S-301 / V-07 / collected 09:00 UTC | Use the independent collection record for the culture timeline; keep measurement at 09:18 separately. |
| Density scale | 4.8 million → 4,800,000 cells/mL | Read the scale in the unit header. Keep density and percentage fields distinct. |
| Preparation evidence | Factor 2 already represented in density | The worked destination retains the confirmed result basis. Do not use a single sample-level multiplier for every analyte. |
| Delivery coverage | 3 entries → 2 analyses + 1 repeated entry | Reconcile identities across snapshots. A new filename is delivery evidence, not evidence of a new scientific result. |
4.8 × 10^6 cells/mL becomes 4,800,000 cells/mL. The confirmed correction state means no further factor of two: 9,600,000 would be wrong for this example. Viability stays 96%, glucose stays 3.2 g/L, and unavailable lactate stays unavailable. The second snapshot adds A-102 while retaining evidence that A-101 was delivered again.
Inspect the source and proposed destination records
Source example
{
"analyzer": "FLEX2-DEMO",
"snapshots": [
{
"file": "handoff-01.csv",
"analyses": [
"A-101"
]
},
{
"file": "handoff-02.csv",
"analyses": [
"A-101",
"A-102"
]
}
],
"selected": {
"analysis": "A-101",
"sample": "S-301",
"vessel": "V-07",
"collectedAt": "2026-09-10T09:00:00Z",
"measuredAt": "2026-09-10T09:18:00Z",
"viableCellDensity": 4.8,
"densityUnit": "10^6 cells/mL",
"viabilityPercent": 96,
"preDilutionMultiplier": 2,
"densityCorrection": "already included",
"glucose": 3.2,
"glucoseUnit": "g/L",
"lactate": null,
"lactateState": "unavailable"
}
}Proposed Seal records
{
"uniqueAnalyses": [
"A-101",
"A-102"
],
"selected": {
"analysis": "A-101",
"sample": "S-301",
"vessel": "V-07",
"cultureTime": "2026-09-10T09:00:00Z",
"measuredAt": "2026-09-10T09:18:00Z",
"viableCellDensity": 4800000,
"densityUnit": "cells/mL",
"viabilityPercent": 96,
"glucose": 3.2,
"glucoseUnit": "g/L",
"lactate": null,
"lactateState": "unavailable"
},
"reconciliation": {
"sourceEntries": 3,
"uniqueAnalyses": 2,
"repeatedEntries": 1
}
}Run these cases against the proposed Seal mapping
These are worked test inputs and expected outcomes to verify in your configured connection.
Two files overlap
Input: Deliver the worked handoffs, then replay handoff-02.csv.
Expected: Two analyses remain. The new arrival is traceable, and its A-101 and A-102 entries resolve to the intended existing results.
The density correction is already included
Input: Provide the worked 4.8 million cells/mL result and factor 2 with explicit already-corrected evidence.
Expected: The destination contains 4,800,000 cells/mL, not 9,600,000. Viability and glucose retain their separately defined quantities.
The source contains no valid lactate number
Input: Use the unavailable lactate entry alongside valid glucose.
Expected: Glucose remains usable under its own rules. Lactate is missing with its source state, and the sample is not presented as a complete numeric analyte panel.
A later evaluation shares the original collection
Input: Deliver a supported reanalysis with a later evaluation time and the same physical collection evidence.
Expected: Both evaluations remain attributable to S-301. The culture timeline still uses the 09:00 collection, and the reviewer can identify the selected evaluation and its supporting evidence.
What to bring to the implementation review
Save the export headers, installed version, identity contract, source reports, collection and preparation evidence, snapshot-completion procedure and three-entry reconciliation. Assign ownership for failed transfers and unresolved results before running unattended collection. Seal then has the context needed for an inspectable culture comparison.
Diagnose the failure from the source evidence
Each snapshot doubles the historical result count
Older rows travel with newer rows in the chosen snapshot design.
Compare the two retained files by source analysis identity and expected row population. In the example, the second import may add A-102; it must resolve A-101 to its existing analysis. Test simultaneous deliveries and an interrupted import.
Density is twice the instrument report
The worked density already contains its preparation correction.
Compare the original report, preparation evidence, exported quantity and mapping calculation. Retain 4.8 million cells/mL as 4,800,000 cells/mL here. Check chemistry and viability independently rather than applying the same multiplier.
A culture appears to have been sampled again
A later evaluation or measurement can be confused with a later physical collection.
Find the source relationship and collection record. Show a reanalysis under its original collection with its own evaluation evidence; create another collection only when physical sampling evidence supports it.
Apply this to your records
Use this with the Nova BioProfile FLEX2 culture results first-run procedure. Retain the actual configuration and evidence, then record what happened in each receiving exercise. Expected behavior is printed as a prompt; your observed result starts blank.
Download the blank working review
Write your working notes here
Notes stay in this tab and are not submitted. Download them before leaving to keep a copy. Use record references to identify your evidence.
Record the installed software, exact export headers, selected Historical Results population and source reports. Describe the owner and mechanism that produce a consistent closed handoff before collection. Identify separately required QC/image evidence.
Match each entry to its analyzer, actual source identity, sample, vessel and independent physical collection. Record repeated-measurement and later-evaluation relationships. For overlapping snapshots, list expected entries, unique analyses and repeated deliveries.
Check density scale, percentage and chemistry units independently. Record preparation evidence and whether each applicable correction is already included. Preserve missing/qualified results and keep collection time separate from analysis/evaluation time.
Input: Deliver the worked handoffs, then replay handoff-02.csv. Expected: Two analyses remain. The new arrival is traceable, and its A-101 and A-102 entries resolve to the intended existing results. Record the actual input/evidence references, observed result, whether it matches the expectation and any unresolved issue.
Input: Provide the worked 4.8 million cells/mL result and factor 2 with explicit already-corrected evidence. Expected: The destination contains 4,800,000 cells/mL, not 9,600,000. Viability and glucose retain their separately defined quantities. Record the actual input/evidence references, observed result, whether it matches the expectation and any unresolved issue.
Input: Use the unavailable lactate entry alongside valid glucose. Expected: Glucose remains usable under its own rules. Lactate is missing with its source state, and the sample is not presented as a complete numeric analyte panel. Record the actual input/evidence references, observed result, whether it matches the expectation and any unresolved issue.
Input: Deliver a supported reanalysis with a later evaluation time and the same physical collection evidence. Expected: Both evaluations remain attributable to S-301. The culture timeline still uses the 09:00 collection, and the reviewer can identify the selected evaluation and its supporting evidence. Record the actual input/evidence references, observed result, whether it matches the expectation and any unresolved issue.
Identify the configuration/mapping revision tested, reviewer, open issues and owner for each next action. Cite the retained source and destination evidence. A filled note is not itself an accepted test; record the review decision through your actual process.
Unanswered sections remain marked ‘Not recorded’ in the download.
Evidence needed to finish
Overlapping handoffs reconcile to the expected unique analysis population, source quantities retain their independent meanings, and culture comparison uses supported collection events. Incomplete copies, missing results and later evaluations remain visible with recovery owners.
Use the connected evidence in a review
Ask neil to compare V-07’s glucose, lactate and cell-density evidence at each recorded collection. Include original values, units, preparation basis and unresolved results. Seal can keep that comparison connected to the experiment without hiding repeats or missing data in an apparently complete trend.
Work through result units and preparation factors →