Skip to content

1. Neutralisation assay

Jump to bottom
Scott Warchal edited this page Feb 14, 2022 · 9 revisions

Assay

For more details see here

Samples

Samples are serum samples from patient's blood which may contain neutralising antibodies.

Neutralisation

Cells (Vero E6) are grown in 384-well plates in the presence of serum samples, each serum sample is assayed at 4 serial dilutions (1:40, 1:160, 1:640, 1:2560) in adjacent wells. Virus of a particular variant is then added to the wells and incubated for 24 hours. After 24 hours they are then fixed with formaldehyde, stained with fluorescent markers, and imaged on the Phenix microscope.

Plate staining & imaging

Plates are stained with 2 fluorescent markers that specifically label cell nuclei or viral plaques, they are then imaged in separate fluorescent channels at 5X magnification with one site per well.

Image analysis

The image analysis is performed in Harmony.

  • Segment cells (DAPI) and viral plaque (488) per well.
  • Measure cell and plaque area per well.
  • Measure cell and plaque intensity per well.

Plate format

Historically the assay was performed in 4 x 96-well plates, with a plate per dilution, and some of the internal code is still structured around this by splitting a 384-well plate up into 4 pseudo 96-well plates. However, the exported assay plates are 384-well plates with the 4 dilutions on a single plate in adjacent wells.

image

i.e wells A01, A02, B01, B02 would be the 4 dilutions of an individual sample.

The assay format is fixed, with sample wells, and control wells: positive control (nanobody) wells; no-virus wells; and virus-only wells.

image

NOTE: Within the plaque_assay codebase wells are typically referred to as their pseudo-96-well labels. They are converted from 384 to 96-well format at the initial ingestion stage.

Barcode format

Plate barcodes are in the format S|T{variant:02}{workflow_id:06} where S is the standard assay, T is the titration assay. variant is encoded as pairs of 2-digit numbers and workflow_id is a zero-padded number.

e.g S01000001 is an assay plate S, has the variant integer 01, and the workflow_id 000001.

  • A workflow_id can contain multiple variants.
  • workflow IDs between assay (S plates) and titration (T plates) are unrelated.

Variant encoding

Each variant is assigned a pair of zero-padded integers in the LIMS system. The first variant (England2) has been assigned [02, 02], the second variant (B117): [03, 04] etc. There are 2 numbers for each variant to denote the 2 replicates for each plate.
i.e S03000500 and S04000500 are a pair of replicate plates for the same workflow_id (500) and variant (B117).

The variant encoding is obtained in the NE_available_strains table in the serology LIMS database and currently looks like so:

mutant_strain plate_id_1 plate_id_2
England2 S01 S02
B117 S03 S04
B1351 S05 S06
FCI_003 S07 S08
B.1.617.2 (India) S09 S10
Crick 19 (Mauritius) S11 S12
B.1.1.529 (Omicron) S13 S14

Note that this table in the wiki will become out of date if (when?) more variants are added.

Exported data

Data is exported from the Phenix to a dedicated directory on camp /camp/project/proj-c19/working/ABNEUTRALISATION/NA_raw_data/.

The exported data is in the format of a directory per 384-well plate, each directory is named {plate_barcode}__{timestamp}-Measurement {n}, e.g S08000304__2021-09-15T18_00_17-Measurement 1.

The directory name usually ends with Measurement 1, but sometimes it may be another number such as Measurement 2, the analysis needs to be flexible enough to handle this possibility.

Within each plate directory, there is a file named indexfile.txt which a tab-delimited file that contains URLs to each image in the plate. There is also a sub-directory named Evaluation{n} where n is typically 1 but may be greater, within the Evaluation{n} directory is a PlateResults.txt file that is a tab-delimited table containing the measurements.

/camp/project/proj-c19/working/ABNEUTRALISATION/NA_raw_data/S08000304__2021-09-15T18_00_17-Measurement 1
├── Evaluation1
│   └── PlateResults.txt
└── indexfile.txt

PlateResults.txt

The PlateResults.txt file has an 8 row header which is ignored: e.g

Database Name   10.6.58.91                                                
Database Location  http://10.6.58.91/ODA/OdaService.asmx                 
Evaluation Signature  8ac22d16-737d-4a9c-b8cf-3d7187b954d5              
Plate Name  S14000578                                                     
Measurement  Measurement 1                                                 
Evaluation  Evaluation1                                                   
                                                                          
[Data]

Then a tab-delimited table:

Row Column Plane Timepoint Viral Plaques (global) - Area of Viral Plaques Area [µm²] - Mean per Well ...

These extracted measurements should not change as they are uploaded to the LIMS database are raw data, and the column names are fixed in the schema.

indexfile.txt

Row Column Plane Timepoint Field Channel ID Channel Name Channel Type URL ImageResolutionX [m] ImageResolutionY [m] ImageSizeX ImageSizeY PositionX [m] PositionY [m] Time Stamp

The indexfile is not used for the analysis with plaque_assay, but it is parsed and uploaded to the LIMS database, and is used in the results viewer to view the well images using the URL column.

Titration data

Data from titration assays is exported to a different directory on CAMP: /camp/project/proj-c19/working/ABNEUTRALISATION/Titration_raw_data/, and the plates have different barcodes starting with T, but otherwise the structure is the same, but with a different plate layout.

Clone this wiki locally