Sequencing-based neutralization assays to estimate serum neutralizing titers in adult and pediatric cohorts using a library of 61 seasonal H3N2 influenza viruses circulating in late 2023 and 17 recent and historical vaccine strains

This repository contains data and analysis of sequencing-based neutralization assays using a library of 78 seasonal H3N2 influenza viruses, including 17 recent and historical vaccine strains (with both egg- and cell-passaging histories) and 61 strains representing circulating H3N2 diversity in November 2023. These experiments were performed by Caroline Kikawa, using method and analysis developed by the Bloom lab and described in Loes et al (2024).

Quick summary

• The actual strains in the library are listed in CSV format in data/H3N2library_2023-2024_allStrains.csv, and the process of library design is described in non-pipeline_analyses/library_design/README.md.

  • The trimmed HA1 sequences are placed in non-pipeline_analyses/library_design/results/2023-2024_H3_library_protein_HA1.fasta
  • The HA ectodomain sequences (with the H3 transmembrane domain removed) are placed in non-pipeline_analyses/library_design/results/2023-2024_H3_library_protein_HA_ectodomain.fasta.
  • The chimeric HA protein construct sequences with upstream signal peptide fixed to WSN sequence, downstream transmembrane domain fixed to H3 consensus, and downstream C-terminal tail fixed to WSN sequence are listed in non-pipeline_analyses/library_design/results/2023-2024_H3_library_protein_constructs.fasta

• This library was then assayed against sera from different human cohorts, specifically:

  • 78 unique human serum samples from the University of Pennsylvania, taken at 0 and 28 days post vaccination (from 39 matched individuals) with the egg-based 2022-2023 seasonal influenza virus vaccine. The H3N2 vaccine component that season was A/Darwin/9/2021. Metadata for these sera are placed in data/sera_metadata/metadata_PennVaccineCohort.csv
  • 56 pediatric human serum samples from Seattle Children's Hospital obtained from routine blood draws. Metadata for these sera are placed indata/sera_metadata/metadata_SCH.csv
  • 20 sera samples from a cohort from Australia taken pre- and post-vaccination from 10 individuals vaccinated with the cell-based 2024 Southern Hemisphere vaccine. Metadata for these sera are placed in data/sera_metadata/metadata_AusVaccineCohort.csv
  • Some pools of the sera from the different cohorts.

• The aggregated titers for each cohort are in in results/aggregated_titers. Specifically:

  • titers for the Penn vaccine cohort: results/aggregated_titers/titers_PennVaccineCohort.csv
  • titers for the Seattle Children's Hospital cohort: in results/aggregated_titers/titers_SCH.csv
  • titers for the Australia vaccine cohort: results/aggregated_titers/titers_AusVaccineCohort.csv
  • titers for the pooled sera: results/aggregated_titers/titers_PooledSera.csv

Description of input data

The input data are in data/, and include:

• data/miscellaneous_plates/: directory of CSV files detailing well IDs of plates requiring barcoding counting but do not need curves fit. This is useful for library titrations to estimate strain balancing and titers. See seqneut-pipeline documentation on miscellaneous_plates for more details.
• data/neut_standard_sets/: directory containing the CSV listing the barcodes linked to the spike-in non-neutralized control RNAs. See Loes et al. 2024 for more details.
• data/plates/: directory of CSV files that maps 96-well plate well IDs to each individual serum-containing or no-serum well. See seqneut-pipeline documentation on plates for more details.
• data/viral_libraries/: directory of CSV files that match 16-nucleotide barcodes to the library strain names.
• data/H3N2library_2023-2024_allStrains.csv: CSV of strain names in the library that dictates the order viruses are plotted on the X-axis of aggreagted NT50 plots.

Library background and design

Information on the library design is placed in non-pipeline_analyses/library_design/.

Library pooling and quality checks

Analysis of library strain balancing and calculations for re-pooling library are placed in non-pipeline_analyses/library_pooling/. These are currently manually run notebooks on barcode counts processed from miscellaneous_plates.

Running the pipeline

This repository contains an analysis of the data using the Bloom lab software seqneut-pipeline as a submodule. See that repository for intstructions on how to use Github submodules, including seqneut-pipeline.

The configuration for the analysis is in config.yml and the analysis itself is run by snakemake using Snakefile. Again, see seqneut-pipeline for more description of how the pipeline works.

To run the pipeline, build the seqneut-pipeline conda environment from the environment.yml in seqneut-pipeline. Then run the pipeline using:

snakemake -j <n_jobs> --software-deployment-method conda

To run on the Hutch cluster, you can use the Bash script run_Hutch_cluster.bash.

Exploring the results

The results are placed in results/, and HTML rendering of results is in docs/. For interactive plots, see the GitHub pages at https://jbloomlab.github.io/flu_seqneut_H3N2_2023-2024/

Running the pipeline will generate many curves, which are stored as 'pickle' objects. To generate custom plots with specific curves, use the manually run notebook in notebooks/plot_specific_sera.py.ipynb. There are also detailed analysis notebooks generated by this pipeline. The final titers for each serum-virus pair are in results/aggregated_titers/ The results/ also contains additional more detailed files with the per-viral-barcode counts for each sample, the fraction infectivities used to fit the neutralization curves, and details on the curve fits and titers for each individual replicate.

There are also custom plotting notebooks in non-pipeline_analyses/, which are works-in-progress.