Repository for the code used to generate figures for Zhou et al. and running the BEAN-FUSE pipeline.
crispr-BEAN-FUSE is a submodule that can be used to run the BEAN-FUSE pipeline. Please see the following links to learn more about the BEAN and FUSE pipeline respectively.
Note that the code and data used to generate Figure 4 or S4 are not made available on this Github as these figures use confidential data from the UK Biobank. Controlled access, patient-level data from the UK Biobank may be requested at https://ams.ukbiobank.ac.uk/ams/.
Figure_Plots_Zhou_et_al contains the data, scripts, and outputs relevant to rendering all of the figures in the paper, which are located in the directories data, scripts, and output respectively.
ldlr_fl_analysis contains scripts for analyzing LDLR-FL raw reporter data (.fastq files). reporter_information_reference.csv is the input file for analysis, editing_analysis_version1.py is for analyzing shorter NGS reads, and editing_analysis_version2.py is for analyzing longer reads.
Description of figures:
Figure 1. Activity-normalized prime editing screening (ANPE) pipeline and benchmarking.
Figure 2. LDLR137-219 screen functional scores reflect domain conservation and computational pathogenicity scores.
Figure 3. Analysis of correlates of prime editing efficiency and functional scores from LDLR-FL screen.
Figure 4. LDLR prime editing functional scores enhance clinical variant interpretation.
Figure 5. LDLR class A repeat 5 harbors LDL uptake-increasing variants that functionally complement pathogenic variants.
Figure 6. Strong LDLR GOF variants generate novel LDLR-APOB inter-atom contacts.
Figure 7. Comparison of LDLR variant effect data from CRISPR screening and cDNA deep mutational scanning.
Figure S1. LDLR prime editing library design.
Figure S2. Benchmarking the ANPE pipeline with LDLR137-219.
Figure S3. Assessing reproducibility, and concordance, and off-target effects of LDLR-FL data.
Figure S4. Comparison of LDLR-FL FUSE data to UK BioBank LDL-C levels.
Figure S5. In silico stability and affinity analysis of LDLR variants.
Figure S6. LDLR-mCherry expression and LDL binding versus uptake tests.
Figure S7. Variant effect characterization on LDLR splicing.
git clone https://github.com/pzhou1729/zhou-et-al.git
Zhou, P., Velimirovic, M., Yu, T. et al. LDLR variant classification through activity-normalized prime editing screening. bioRxiv (2025). https://www.biorxiv.org/content/10.64898/2025.12.16.694467v1.abstract
Questions and PRs welcome — please open an issue.