This repository contains code for

• CSV parsers for the Hamilton logs and qPCR results from the QuantStudio7 Design and Analysis software (server.js)
• API calls to eLab (server.js)
• Example Hamilton logs (example-hamilton-logs)
• Example QuantStudio results (example-qpcr-logs)

Installation Requirements

Because the servers have limited access to the internet, steps 1 & 2 should occur on a separate machine

1. Download the LTS version of NodeJS
2. Run npm install on the CLI in this directory (where package.json is) to download dependencies
3. Copy the entire directory (including node_modules) to the server
4. Copy the node executable to the server
5. Configure a link to the executable so that the keyword node is recognized
6. Type node --version on the CLI to verify
7. Enter in eLabs login token in the the config files in /config
8. The NODE_ENV variable must be set which matches one of the config file names. For example, export NODE_ENV=sandbox

Logging

• Daily error logs will print to logs/error-M-DD-YYYY.log
  • Will log a new line for every file that is processed, regardless if error occurred. This is for email parsing purposes.
• Daily combined logs (both errors and info) will print to logs/combined-M-DD-YYYY.log
• The timestamp in each file will default to the timezone of the server, and are formatted as M/D/YYYY, hh:mm:ss A Z, where Z is the offset from UTC
  • For example, a server configured to EST will be displayed as 9/1/2020, 11:04:24 AM -04:00

Processing Requirements

• Hamilton output format
  • We expect 2 empty lines at the start of the file
  • Expected headers:
    • Protocol - must be SAMPLE_ALIQUOT, RNA_EXTRACTION, or QPCR_PREP
    • Sample Tube Barcode
    • Output Barcode
    • Output Well Number
    • UserName
    • Machine SN
    • Reagent Names - strings in this file must match exactly the field it maps to on eLabs
    • Reagent Lot Numbers - length of this array must match length of reagent names
    • eLab Sample ID - can be blank
• QuantStudio output format
  • We expect 2 additional lines between the comments and headers
  • For Ct values, we expect the Target Call file to have the headers
    • Well Position
    • Sample
    • Target
    • Cq
  • For Call values, we expect the Well Call file to have the headers
    • Well Position
    • Sample
    • Call - valid values are WARNING, POSITIVE, NEGATIVE, INCONCLUSIVE, or INVALID
    • It must NOT contain the header Cq
    • We expect the file to contain the string [Well Call]
    • We expect the serial number to follow the string # Instrument Serial Number:
    • We expect the technician's initials to follow the string # User Name:
  • Only the Well Call file updates the Sample Process Status, qPCR User, and qPCR SN.
  • All samples named PCR_POS, NEC_#, NTC_# is assumed to be a control and skipped during processing UNLESS there is a failure. In which case, the Control Fail Procedure will be followed, outlined below.
• Aliquot (Gandalf) prefetch file
  • We expect the input file to have 2 empty lines at the start of the file
  • Input file expected headers:
    • Sample Tube Barcode
  • The input file will be overwritten and will not contain empty lines at the start of the file
  • Output file headers:
    • Sample Tube Barcode
    • eLab Sample ID - either the IDs or error values: NOT FOUND, DUPLICATE, INDIVIDUAL POOLED, or ERROR
• Existing eLab SampleType and custom fields should not be altered
  • Expected fields are defined in constants.js
• File contents and eLab fields are case and white-space sensitive

Control Fail Procedure

Control samples in every 384 qPCR plate are expected to have the same layout. Control samples must go onto the plate according to the following figure and A2, B1, B2 are empty. The rest of plate are patient samples.

     1    |    2
A PCR_POS |  /
-------------------
B    /    |  /
-------------------
C  NTC_1  |  NTC_3
-------------------
D  NTC_2  |  NTC_4
-------------------
E  NEC_1  |  NEC_3
-------------------
F  NEC_2  |  NEC_4

• Control sample failures will show WARNING in qPCR results
  • If PCR Positive Control fails, all 4 96 well plates need to be re-prepped for qPCR.
  • IF NEC or NTC fails, all samples from the failed plate(s) need to have RNA re-extracted.
  • If WARNING is found outside of these 9 wells, no samples will be processed.
• See qPCR Plate Layout to match patient samples to control sample.
• Re-extraction takes precedence.
  • For example, if PCR_POS and NTC_1 both fail, Plate 1 needs to be re-extracted, and the other 3 plates re-prepped.

Any samples that fail 5x in this process are considered invalid and need to be recollected. Status is to be considered Finished.

qPCR Plate Layout

Stamping configuration of 4 qPCR Prep plates onto one qPCR plate

Key: Plate 1 (Red), Plate 2 (Green), Plate 3 (Yellow), Plate 4 (Blue)

Pooled Sample Processing

• All samples in SAMPLE_ALIQUOT and RNA_EXTRACTION logfiles are expected to be parent "Pooled COVID-19 Sample" or "COVID-19 Sample" performed as "Individual". No children are updated in these 2 processes.
• Children performed as "Pooled" will be updated from QPCR_PREP, Well Call, and Target Call files
  • Children performed as "Individual" will be silently skipped
• Children whose parent has an invalid test result will be marked as "qPCR Completed" & "Invalid" & "Individual"
• Children whose parent has a positive test result will be marked as "qPCR Completed" & "Presumptive positive" & "Individual"
• Pooled COVID-19 Sample will always be set to "qPCR Completed", never "Finished"