adjing

Scripts to regenerate all analyses and figures presented in:

Pre-treatment T-cell Transcriptional Signatures Predict Immunotherapy Outcomes in Melanoma

Noah Lepola^1,2, Caroline Dravillas³, Shannon Gray^1,2, Michael S. Bodnar^1,2, Namrata Arya^1,2, Richard Wu³, Claire Verschraegen³, William E Carson³, Kari L Kendra³, Daniel J Spakowicz^{3,4^}, Christin E Burd^{1,2,4^
1} Department of Molecular Genetics, The Ohio State University College of Arts and Sciences, Columbus, Ohio
² Department of Cancer Biology and Genetics, The Ohio State University College of Medicine, Columbus, Ohio
³ Division of Medical Oncology, Department of Internal Medicine, The Ohio State University Comprehensive Cancer Center; Columbus, OH, USA.
⁴ Pelotonia Institute for Immuno-Oncology, The Ohio State University Comprehensive Cancer Center; Columbus, OH, USA.

Preprint: Research Square (doi:10.21203/rs.3.rs-10629841/v1)

Submitted for publication.

Graphical abstract

Repository structure

Path	Contents
man/figures/	Hex sticker (logo.png) and graphical abstract used in this README
manuscript/	Scripts, shared helpers, and outputs used to regenerate the manuscript figures and supporting tables
manuscript/scripts/	R Markdown notebooks that process data and create each figure panel
manuscript/figures/	PNG figure panels written by the scripts above
manuscript/Outputs/	Intermediate R objects (.RData / .rds), model fits, and derived tables consumed by the figure scripts
manuscript/rf_functions.R	Shared random-forest and plotting helpers sourced by the figure notebooks
exploratory/	Earlier exploratory analyses not used for the final manuscript figures
00-paths.R / adjing.json	Local path configuration for clinical and NanoString source data
adjing.Rproj	RStudio project file

The manuscript/ directory is the entry point for regenerating the published analyses. Numbered notebooks (00_, 01_, 02_) prepare adjuvant and metastatic cohort objects; subsequent notebooks train models and write figure panels into manuscript/figures/. Script filenames encode the figure panels they produce (for example, adj_rec_Fig2ABCDGH.Rmd generates Figure 2 panels A–D, G, and H).

Figure-to-script map

Use this table to locate the notebook that regenerates a given panel. Output PNGs are written under manuscript/figures/ unless noted otherwise.

Main figures

Figure panels	Script	Output files
Fig. 2A, 2B, 2C, 2D, 2G, 2H	adj_rec_Fig2ABCDGH.Rmd	AdjRecLogr.png (2A), AdjRecClinical.png (2B), AdjRecFullpanel.png (2C), AdjRecBoruta.png (2D), AdjRecDistribution.png (2H)
Fig. 2E	logr_adj_rec_and_volcano_Fig_2E.Rmd	RecAdjVol.png
Fig. 2F, Fig. 3D	F1_opt_adj_Fig2F_Fig3D.Rmd	RecAdjOptBar.png (2F), ToxAdjOptBar.png (3D)
Fig. 3A, 3B, 3E	adj_tox_Fig3ABE.Rmd	AdjToxDualGraph.png (3A), AdjToxBoruta.png (3B), AdjToxLogr.png (3E)
Fig. 3C	logr_adj_tox_and_volcano_Fig_3C.Rmd	ToxAdjVol.png
Fig. 5A, 5B, 5E	meta_pro_Fig5ABE.Rmd	MetaProDualGraph.png (5A), MetaProBoruta.png (5B), MetaProLogr.png (5E)
Fig. 5C	logr_meta_pro_and_volcano_Fig_5C.Rmd	ProMetVol.png
Fig. 5D, Fig. S1D	F1_opt_meta_Fig5D_FigS1D.Rmd	ProMetOptBar.png (5D), ToxMetOptBar.png (S1D)
Fig. 6A, 6B; Fig. S2A, S2B	cross_setting_comparison_Fig6AB_FigS2AB.Rmd	AdjTrainRecCrossCohort.png (6A), MetaTrainRecCrossCohort.png (6B), AdjTrainToxCrossCohort.png (S2A), MetaTrainToxCrossCohort.png (S2B)

Supplementary figures

Figure panels	Script	Output files
Fig. S1A, S1B, S1E	meta_tox_FigS1ABE.Rmd	MetaToxDualGraph.png (S1A), MetaToxBoruta.png (S1B), MetaToxLogr.png (S1E)
Fig. S1C	logr_meta_tox_and_volcano_Fig_S1C.Rmd	ToxMetVol.png
Fig. S1D	F1_opt_meta_Fig5D_FigS1D.Rmd	ToxMetOptBar.png
Fig. S2A, S2B	cross_setting_comparison_Fig6AB_FigS2AB.Rmd	AdjTrainToxCrossCohort.png (S2A), MetaTrainToxCrossCohort.png (S2B)

Data preparation and supporting tables

These notebooks do not write main-figure panels, but they create the cohort objects, clinical tables, and intermediate results used by the figure scripts:

Script	Role
00_transcript_processing_adjuvant.Rmd	Process adjuvant NanoString / clinical characteristics
01_outcome_processing_adjuvant.Rmd	Process adjuvant patient outcomes
02_object_processing_metastatic.Rmd	Build metastatic cohort clinical and outcome objects
tableone_adjuvant.Rmd	Adjuvant cohort Table 1
tableone_metatatic.Rmd	Metastatic cohort Table 1
boruta_transcript_counts.Rmd	Boruta feature-selection transcript counts across outcomes
[adjing timeline.Rmd](manuscript/scripts/adjing timeline.Rmd)	Adjuvant cohort timelines (manuscript/Outputs/)
[metastatic timeline.Rmd](manuscript/scripts/metastatic timeline.Rmd)	Metastatic cohort timelines (manuscript/Outputs/)

To regenerate a figure, open the corresponding notebook under manuscript/scripts/, ensure paths in adjing.json point to the source data on your system, knit the upstream processing notebooks if manuscript/Outputs/ objects are not already present, then knit the figure notebook.