Last Updated on 2025-03-28
This repository contains the supporting figures and tables for: Comparative analysis of transcriptomic and proteomic expression between two non-small cell lung cancer subtypes: https://doi.org/10.1021/acs.jproteome.4c00773
The supporting tables are csv files in the tables folder. The column names and contents of the csv files
are described in Tables 1-6 below.
- Figures S1 and S2: PCA bi-plots of NSCLC subtypes PBMC and NAT comparisons
- Figures S3 and S4: Volcano plots of NSCLC subtypes PBMC and NAT comparisons.
- Figures S5 to S7: Heatmaps of NSCLC DEGs and DEPs.
- Figure S8: Bar plots of functional enrichment between NSCLC subtypes and PBMC and NAT.
- Gene counts from the HISAT2 alignments estimated by featureCounts.
- Gene counts from transcript classification by Salmon.
- Differential gene expression edgeR outputs.
- Peaks normalised top 3 peptide intensities.
- Differential protein expression DEqMS outputs.
- Functional enrichment analysis g:Profiler outputs.
Figure S1: Bi-plots of the LUSC and PBMC and NAT comparison. (A) PCA of normalised gene count matrix numbered with donor identifier. LUSC (blue) & PBMC (grey). (B) PCA of normalised gene count matrix with the genes contributing to the PC directions annotated. (C) PCA of normalised top 3 peptide intensities numbered with donor identifier. LUSC (light blue) & NAT (grey). (D) PCA of normalised top 3 peptide intensities with the protein contributing to the PC directions annotated.
Figure S2: Bi-plots of the LUAD and PBMC and NAT comparison. (A) PCA of normalised gene count matrix numbered with donor identifier. LUAD (red) & PBMC (grey). (B) PCA of normalised gene count matrix with the genes contributing to the PC directions annotated. (C) PCA of normalised top 3 peptide intensities numbered with donor identifier. LUAD (purple) & NAT (grey). (D) PCA of normalised top 3 peptide intensities with the protein contributing to the PC directions annotated.
Figure S3: (A) Comparison of LUSC & PBMC (n=17,719). Thresholds are represented by dotted lines at FDR of 1% and log2 fold change of 1.5. (B) Comparison of LUSC & NAT (n=1,330). Thresholds are represented by dotted lines at p-value of 1% and log2 fold change of 1.
Figure S4: (A) Comparison of LUAD & PBMC (n=17,586). Thresholds are represented by dotted lines at FDR of 1% and log2 fold change of 1.5. (B) Comparison of LUAD & NAT (n=1,478). Thresholds are represented by dotted lines at p-value of 1% and log2 fold change of 1.
Figure S5: (A) Comparison of LUSC & LUAD DEGs below a FDR of 1%. (n=428). (B) Comparison of LUSC & LUAD DEPs below a p-value 1% (n=139). Colour bar shows log2 fold change rescaled as z-scores i.e. each unit from zero represents one standard deviation from the row average value for each protein.
Figure S6: (A) Comparison of LUSC & PBMC DEGs below a FDR of 1%. (n=8,089). (B) Comparison of LUSC & NAT DEPs below a p-value 1% (n=379). Colour bar shows log2 fold change rescaled as z-scores i.e. each unit from zero represents one standard deviation from the row average value for each protein.
Figure S7: (A) Comparison of LUAD & PBMC DEGs below a FDR of 1%. (n=10,058). (B) Comparison of LUAD & NAT DEPs below a p-value 1% (n=234). Colour bar shows log2 fold change rescaled as z-scores i.e. each unit from zero represents one standard deviation from the row average value for each protein.
Figure S8: Bar plots of functional enrichment between NSCLC subtypes and PBMC and NAT. Statistical significance level indicated by the -log10 p-value on the x-axis. (A) GO biological processes enriched in NSCLC subtypes. (B) Reactome pathways enriched in NSCLC subtypes.
Transcripts were quantified by genomic alignments using HISAT2 (version 2.2.1) [1] and featureCounts (version 2.0.6) [2], and by transcript classification using Salmon (version 1.10.3) [3].
Tables S1-3 contain the gene counts from the HISAT2 alignments estimated by featureCounts.
Table 1
File
──────────────────────────────────────────
Table-S1-Hisat-LUAD-vs-PBMC-counts.csv
Table-S2-Hisat-LUSC-vs-PBMC-counts.csv
Table-S3-Hisat-LUSC-vs-LUAD-counts.csv
──────────────────────────────────────────
Column names: File
| Column name | Description |
|---|---|
name |
Ensembl gene identifier |
gene |
HGNC gene symbol |
sample_id the donor id or donor id suffixed with T for tumour or N for PBMC samples. |
mapped read counts from featureCounts |
HISAT2 Counts Tables Information
Tables S4-6 contain the gene counts from transcript classification by Salmon.
Table 2
File
───────────────────────────────────────────
Table-S4-Salmon-LUAD-vs-PBMC-counts.csv
Table-S5-Salmon-LUSC-vs-PBMC-counts.csv
Table-S6-Salmon-LUSC-vs-LUAD-counts.csv
───────────────────────────────────────────
Column names: File
| Column name | Description |
|---|---|
name |
Ensembl gene identifier |
gene |
HGNC gene symbol or Ensembl if missing |
length |
the length of the target transcript |
sample_id the donor id or donor id suffixed with T for tumour or N for PBMC samples. |
mapped reads counts by Salmon |
Salmon Counts Table Information
Differential gene expression (DEG) was estimated using EdgeR and default settings [4]. Results were filtered for common DEG from both HISAT2 and Salmon counts
Tables S7-9 contain the edgeR outputs.
Table 3
File
───────────────────────────────────────
Table-S7-edgeR-DEG-LUAD-vs-PBMC.csv
Table-S8-edgeR-DEG-LUSC-vs-PBMC.csv
Table-S9-edgeR-DEG-LUSC-vs-LUAD.csv
───────────────────────────────────────
Column names: File
| Column name | Description |
|---|---|
name |
Ensembl gene identifier |
gene |
HGNC gene symbol or Ensembl if missing |
baseMean |
mean read counts |
baseMeanA |
mean read count group A |
baseMeanB |
mean read count group B |
foldChange |
fold change B/A |
log2FoldChange |
log2 fold change B/A |
PValue |
p-value |
PAdj |
Benjamini-Hochbergadjusted p-value |
FDR |
False discovery rate |
falsePos |
false discovery counts |
sample_id the donor id or donor id suffixed with T for tumour or N for PBMC samples. |
sample HISAT2 read count |
edgeR Table information
Label free quantification using the Peaks Q module of Peaks Studio [5,6] yielding matrices of protein identifications as quantified by their normalised top 3 peptide intensities.
Tables S10-12 contain normalised top 3 peptide intensities.
Table 4
File
──────────────────────────────────────────────────────────────
Table-S10-Peaks-top3-peptides-intensities-LUAD-vs-NAT.csv
Table-S11-Peaks-top3-peptides-intensities-LUSC-vs-NAT.csv
Table-S12-Peaks-top3-peptides-intensities-LUSC-vs-LUAD.csv
──────────────────────────────────────────────────────────────
Column names: File
| Column name | Description |
|---|---|
protein |
protein short name |
gene |
HGNC gene symbol |
sample_id the donor id or donor id suffixed with T for tumour or N for NAT samples |
Normalised top 3 peptide intensity from Peaks |
Peaks normalised Top 3 peptide intensities Table information
The normalised top 3 peptide intensities were filtered to remove any proteins for which there were more than two missing values across the samples. Differential protein expression (DEP) was then calculated with DEqMS using the default steps [7].
Tables S13-15 contain the outputs of DEqMS.
Table 5
File
────────────────────────────────────────
Table-S13-DEqMS-DEP-LUAD-vs-NAT.csv
Table-S14-DEqMS-DEP-LUSC-vs-NAT.csv
Table-S15-DEqMS-DEP-LUSC-vs-LUAD.csv
────────────────────────────────────────
Column names: File
| Column name | Description |
|---|---|
logFC |
log2 fold change between two groups |
AveExpr |
the mean of the log2 ratios across all samples |
t |
Limma t-values |
P.Value |
Limma p-values |
adj.P.Val |
BH method adjusted Limma p-values |
B |
Limma B values |
gene |
HGNC gene symbol |
count |
peptide count values |
sca.t |
DEqMS t-statistics |
sca.P.Value |
DEqMS p-values |
sca.adj.pval |
BH method adjusted DEqMS p-values |
protein |
protein short name |
DEqMS Table information
Functional enrichment analysis used g:Profiler [8] using G:OSt multi_query in default settings for homo sapiens modified to exclude GO electronic annotations. Gene ids were used as inputs for DEGs and protein ids for DEPs.
We used four lists as inputs to the multiple query setting for G:OSt for comparing NSCLC subtypes for DEGs filtered at thresholds 5% FDR and logFC 1.5 and DEPs filtered p-val 5% and logFC 1 for protein expression. The identifiers for each list are LUAD Genes: LUAD DEGs, LUAD Proteins: LUAD DEPs, LUSC Genes : LUSC DEGs, LUSC Proteins: LUSC DEPs.
The four lists for comparison of NSCLC subtypes with PBMC were NSCLC DEGs filtered at 1% FDR and logFC 1.5 and NSCLC DEPs 5% FDR and logFC of 1 for NAT comparison. The list identifiers are as for NSCLC subtype comparison.
Table S16 contain the g:Profiler outputs for the NSCLC comparisons and Table S17 the NSCLC and PBMC/NAT comparisons.
Table 6
File
────────────────────────────────────────────
Table-S16-gprofiler-DEG-LUSC-vs-LUAD.csv
Table-S17-gprofiler-DEG-NSCLC-vs-NORM.csv
────────────────────────────────────────────
Column names: File
| Column name | Description |
|---|---|
term_id |
unique term identifier |
p_values |
hypergeometric p-value after correction for multiple testing |
significant |
indicator for statistically significant results |
term_size |
number of genes that are annotated to the term |
query_sizes |
number of genes that were included in the query |
intersection_sizes |
the number of genes in the input query that are annotated to the corresponding term |
source |
the abbreviation of the data source for the term (e.g. GO:BP) |
term_name |
the ontology term name |
effective_domain_size |
the total number of genes "in the universe" used for the hypergeometric test |
source_order |
numeric order for the term within its data source |
parents |
list of term IDs that are hierarchically directly above the term. For non-hierarchical data sources this points to an artificial root node. |
id |
The identifier of for the input list associated with the row e.g. LUAD Genes were the list of LUAD DEGs. |
g:Profiler Table information
1. Kim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nature Biotechnology. 2019;37: 907–915. doi:10.1038/s41587-019-0201-4
2. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2013;30: 923–930. doi:10.1093/bioinformatics/btt656
3. Srivastava A, Malik L, Sarkar H, Patro R. A Bayesian framework for inter-cellular information sharing improves dscRNA-seq quantification. Bioinformatics. 2020;36: i292–i299. doi:10.1093/bioinformatics/btaa450
4. Yunshun Chen , Aaron Lun, Davis McCarthy , Xiaobei Zhou , Mark Robinson, Gordon Smyth. edgeR. 2017. doi:10.18129/B9.BIOC.EDGER
5. Zhang J, Xin L, Shan B, Chen W, Xie M, Yuen D, et al. PEAKS DB: De novo sequencing assisted database search for sensitive and accurate peptide identification. Molecular & Cellular Proteomics. 2012;11: M111010587.
6. Lin H, He L, Ma B. A combinatorial approach to the peptide feature matching problem for label-free quantification. Bioinformatics. 2013;29: 1768–1775. doi:10.1093/bioinformatics/btt274
7. DEqMS. Available: http://bioconductor.org/packages/DEqMS/
8. Kolberg L, Raudvere U, Kuzmin I, Vilo J, Peterson H. gprofiler2 – an R package for gene list functional enrichment analysis and namespace conversion toolset g:Profiler. F1000Research. 2020;9: 709. doi:10.12688/f1000research.24956.2