Characterizing the epitope of a public antibody clonotype to SARS-CoV-2 spike HR1 using deep mutational scanning

This README describes the analysis in:
Evidence of antigenic drift in the fusion machinery core of SARS-CoV-2 spike

Input files:

• Raw reads from BioProject PRJNA888135
• ./data/barcodes.tsv: Internal barcodes for NGS error correction
• ./Fasta/HR1_ref_seq.fa: Reference (i.e. wild type) amino acid sequences
• ./data/S2HR1_exp_fus_scores.tsv: Expression and fusion scores from a previous study

Computing binding scores from raw Illumina MiSeq reads

1. Use PEAR to merge forward and reverse reads.
2. Run python script/S2HR1_bind_fastq2count.py to generate sequence count file.
3. Run python script/S2HR1_bind_count_nuc2aa.py to generate amino acid count file.
4. Run python script/S2HR1_bind_count2score.py to generate binding score file.
5. Run Rscript script/common_muts.R to add expression and fusion scores into the final score file.
6. Run python script/S2HR1_bind_resi_analysis.py to generate mutational effect file
7. Run python script/Bfactor_to_score.py to generate PDB with mutational effect as B factor

Plot correlations between replicates and between antibodies

1. Run Rscript script/plot_correlations.R to compare binding scores across antibodies and replicates.
2. Run Rscript script/plot_QC.R to compare binding scores among nonsense, missense, and silent mutations.
3. Run Rscript script/plot_bind_vs_exp to compare binding and expression scores.

Analyzing mutational effects

1. Run Rscript script/plot_adj_bind_score.R to plot mutational effects along residue positions.
2. Run pymol script/plot_Bfactor_as_score.pml to generate structural figure colored according to mutational effects.

Expeirmental validation results

• See ./flow_cytometry