2 |
3 |
4 | 
5 |
6 |
7 |
8 | ## Overview
9 |
10 | PROGENy is resource that leverages a large compendium of publicly available
11 | signaling perturbation experiments to yield a common core of pathway responsive
12 | genes for human and mouse. These, coupled with any statistical method, can be
13 | used to infer pathway activities from bulk or single-cell transcriptomics.
14 |
15 | This is an R package for storing the pathway signatures. To infer pathway
16 | activities, please check out
17 | [decoupleR](https://doi.org/10.1093/bioadv/vbac016), available in
18 | [R](https://saezlab.github.io/decoupleR/) or
19 | [python](https://github.com/saezlab/decoupler-py).
20 |
21 | ## Installation
22 |
23 | Progeny is available in
24 | [Bioconductor](https://www.bioconductor.org/packages/release/bioc/html/progeny.html).
25 | In addition, one can install the development version from the Github repository:
26 |
27 | ```r
28 | ## To install the package from Bioconductor
29 | if (!requireNamespace("BiocManager", quietly = TRUE))
30 | install.packages("BiocManager")
31 |
32 | BiocManager::install("progeny")
33 |
34 | ## To install the development version from the Github repo:
35 | devtools::install_github("saezlab/progeny")
36 | ```
37 |
38 | ## Updates
39 |
40 | Since the original release, we have implemented some extensions in PROGENy:
41 |
42 | 1. **Extension to mouse**:
43 | Originally PROGENy was developed for the application to human data.
44 | In a benchmark study we showed that PROGENy is also applicable to mouse data,
45 | as described in
46 | [Holland et al., 2019](https://doi.org/10.1016/j.bbagrm.2019.194431).
47 | Accordingly, we included new parameters to run mouse version of PROGENy by
48 | transforming the human genes to their mouse orthologs.
49 | 2. **Expanding Pathway Collection**:
50 | We expanded human and mouse PROGENy with the pathways Androgen, Estrogen and
51 | WNT.
52 | 3. **Extension to single-cell RNA-seq data**:
53 | We showed that PROGENy can be applied to scRNA-seq data, as described in
54 | [Holland et al., 2020](https://doi.org/10.1186/s13059-020-1949-z)
55 |
56 | ## Citation
57 |
58 | > Schubert M, Klinger B, Klünemann M, Sieber A, Uhlitz F, Sauer S, Garnett MJ,
59 | Blüthgen N, Saez-Rodriguez J. 2018. Perturbation-response genes reveal signaling
60 | footprints in cancer gene expression. _Nature Communications_:
61 | [10.1038/s41467-017-02391-6](https://doi.org/10.1038/s41467-017-02391-6)
62 |
--------------------------------------------------------------------------------
/data/model_human_full.rda:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/saezlab/progeny/cad6be0514c3248b9465e48f1cfd2f6a4c3dfb6f/data/model_human_full.rda
--------------------------------------------------------------------------------
/data/model_mouse_full.rda:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/saezlab/progeny/cad6be0514c3248b9465e48f1cfd2f6a4c3dfb6f/data/model_mouse_full.rda
--------------------------------------------------------------------------------
/data/vignette_data.RData:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/saezlab/progeny/cad6be0514c3248b9465e48f1cfd2f6a4c3dfb6f/data/vignette_data.RData
--------------------------------------------------------------------------------
/docs/404.html:
--------------------------------------------------------------------------------
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 | Source: inst/CITATION
M Schubert, B Klinger, M Klünemann, A Sieber, F Uhlitz, S Sauer, MJ Garnett, N Blüthgen, and J Saez-Rodriguez. 2018. “Perturbation-Response Genes Reveal Signaling Footprints in Cancer Gene Expression.” Nature Communications 9 (1)
97 |@Article{,
98 | author = {Michael Schubert and Bertram Klinger and Martina Klünemann and Anja Sieber and Florian Uhlitz and Sascha Sauer and Mathew J Garnett and Nils Blüthgen and Julio Saez-Rodriguez},
99 | title = {Perturbation-response genes reveal signaling footprints in cancer gene expression},
100 | journal = {Nature communications},
101 | year = {2018},
102 | volume = {9},
103 | number = {20},
104 | }
105 | PROGENy is resource that leverages a large compendium of publicly available signaling perturbation experiments to yield a common core of pathway responsive genes for human and mouse. These, coupled with any statistical method, can be used to infer pathway activities from bulk or single-cell transcriptomics.
94 |This is an R package for storing the pathway signatures. To infer pathway activities, please check out decoupleR, available in R or python.
95 |Progeny is available in Bioconductor. In addition, one can install the development version from the Github repository:
100 |
101 | ## To install the package from Bioconductor
102 | if (!requireNamespace("BiocManager", quietly = TRUE))
103 | install.packages("BiocManager")
104 |
105 | BiocManager::install("progeny")
106 |
107 | ## To install the development version from the Github repo:
108 | devtools::install_github("saezlab/progeny")Since the original release, we have implemented some extensions in PROGENy:
114 |128 |130 |Schubert M, Klinger B, Klünemann M, Sieber A, Uhlitz F, Sauer S, Garnett MJ, Blüthgen N, Saez-Rodriguez J. 2018. Perturbation-response genes reveal signaling footprints in cancer gene expression. Nature Communications: 10.1038/s41467-017-02391-6
129 |
Major update with the following main points:
87 |Added the mouse model matrix containing 14 pathways
The human model matrix extended to 14 pathways
Added the following functions: progenyPerm, progenyScatter, progenySavePlots, getModel
Added tests and test data
Added the vignette for usage the PROGENy on single-cell RNA-seq data
Added functionality to work with Seurat objects
This function is designed for getting a model matrix with top significant 73 | genes for each pathway
74 |"Human" or "Mouse" taken from the main function's argument. 85 | Default to "Human"
Desired top number of genes for each pathway according to their 90 | significance(p.value). Default to 100
if TRUE, the model matrix is goign to be generated with a format 95 | makes it directlz compatible with the decoupleR package
getModel()
81 | model_human_full
86 | model_mouse_full
91 | progeny()
96 | progenyPerm()
101 | progenyScatter()
106 | saveProgenyPlots()
111 | vignette_data
116 | R/progenyPermutations.r
66 | progenyPerm.RdCompute progeny pathway scores and assesses significance based on permutations
71 |progenyPerm(
76 | df,
77 | weight_matrix,
78 | k = 10000,
79 | z_scores = TRUE,
80 | get_nulldist = FALSE
81 | )A data.frame of n*m+1 dimension, where n is the number of omic 88 | features to be considered and m is the number of samples/contrasts. 89 | The first column should be the identifiers of the omic features. 90 | These identifiers must be coherent with the identifiers of the weight matrix.
A progeny coefficient matrix. the first column should be 95 | the identifiers of the omic features and should be coherent with 96 | the identifiers provided in df.
The number of permutations to be performed to generate 101 | the null-distribution used to estimate the significance of progeny scores. 102 | The default value is 10000.
if true, the z-scores will be returned for 107 | the pathway activity estimations. Else, the function returns 108 | a normalized z-score value between -1 and 1.
if true, the null score distribution used for 113 | normalization will be returned along with the actual normalized score data 114 | frame.
This function returns a list of two elements. The first element is 122 | a data frame of p*m+1 dimensions, where p is the number of progeny pathways, 123 | and m is the number of samples/contrasts. Each cell represents the 124 | significance of a progeny pathway score for one sample/contrast. The 125 | significance ranges between -1 and 1. The significance is equal to x*2-1, x 126 | being the quantile of the progeny pathway score with respect to the null 127 | distribution. Thus, this significance can be interpreted as the equivalent of 128 | 1-p.value two-sided test over an empirical distribution) with the sign 129 | indicating the direction of the regulation. The second element is the null 130 | distribution list (a null distribution is generated for each sample/contrast).
131 |# use example gene expression matrix
136 | gene_expression <- as.matrix(read.csv(system.file("extdata",
137 | "human_input.csv", package = "progeny"), row.names = 1))
138 |
139 | # calculate pathway activities
140 | progeny(gene_expression, scale=TRUE, organism="Human", top=100, perm=10000)
141 | #> Androgen EGFR Estrogen Hypoxia JAK.STAT MAPK NFkB p53
142 | #> SRR1039508 -0.8380 0.1996 -0.3470 0.9996 0.7718 0.4066 -0.5552 0.9240
143 | #> SRR1039509 -0.4918 0.0608 -0.6360 0.9998 0.4858 0.3284 -0.4426 0.8778
144 | #> SRR1039512 -0.7836 0.2158 -0.4054 0.9996 0.6992 0.4132 -0.4322 0.9230
145 | #> SRR1039513 -0.4372 0.0790 -0.6722 1.0000 0.6724 0.4352 -0.4490 0.7916
146 | #> SRR1039516 -0.7578 0.2614 -0.2944 0.9992 0.6132 0.4200 -0.3042 0.9276
147 | #> SRR1039517 -0.5104 0.5540 -0.6566 1.0000 0.4672 0.7292 -0.3776 0.8090
148 | #> SRR1039520 -0.8056 0.3332 -0.3886 0.9994 0.6094 0.5606 -0.6314 0.9022
149 | #> SRR1039521 -0.3562 0.1490 -0.6650 0.9994 0.4996 0.5776 -0.6296 0.7778
150 | #> PI3K TGFb TNFa Trail VEGF WNT
151 | #> SRR1039508 0.1660 0.5416 -0.5988 -0.3570 0.0432 -0.1070
152 | #> SRR1039509 0.0984 0.7620 -0.4628 -0.3012 0.0508 -0.1712
153 | #> SRR1039512 0.1418 0.6930 -0.4706 -0.3462 0.0842 -0.1586
154 | #> SRR1039513 -0.0706 0.8240 -0.4828 -0.2668 0.0958 -0.3054
155 | #> SRR1039516 0.1706 0.6926 -0.3564 -0.3356 0.2734 -0.3268
156 | #> SRR1039517 0.3604 0.8662 -0.4248 -0.3586 0.2530 -0.4108
157 | #> SRR1039520 0.1788 0.6872 -0.6702 -0.2790 -0.0824 -0.1208
158 | #> SRR1039521 0.0350 0.8092 -0.6476 -0.2630 -0.0724 -0.3158
159 | List with three elements: the gene counts, the experimental design and 73 | the result of limma differential analysis
74 |List with three elements: the gene counts, the experimental design and 80 | the result of limma differential analysis
gene counts
experiemental design
differential analysis result using limma
https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE119931
94 |data("vignette_data")
99 |