diff --git a/README.Rmd b/README.Rmd index 863a76b3..7a0aae20 100755 --- a/README.Rmd +++ b/README.Rmd @@ -107,8 +107,8 @@ my_theme = ) utils::download.file("https://zenodo.org/records/11201167/files/counts_SE.rda?download=1", destfile = "counts_SE.rda") -counts_SE = here("counts_SE.rda") |> load() -tibble_counts = counts_SE %>% as_tibble() +here("counts_SE.rda") |> load() +tibble_counts = counts_SE |> as_tibble() ``` @@ -142,7 +142,7 @@ class(counts_SE) First of all, you can cite all articles utilised within your workflow automatically from any tidybulk tibble ```{r eval=FALSE} -counts_SE %>% get_bibliography() +counts_SE |> get_bibliography() ``` ## Aggregate duplicated `transcripts` @@ -153,7 +153,7 @@ tidybulk provide the `aggregate_duplicates` function to aggregate duplicated tra TidyTranscriptomics ```{r aggregate, message=FALSE, warning=FALSE, results='hide', class.source='yellow'} rowData(counts_SE)$gene_name = rownames(counts_SE) -counts_SE.aggr = counts_SE %>% aggregate_duplicates(.transcript = gene_name) +counts_SE.aggr = counts_SE |> aggregate_duplicates(.transcript = gene_name) ```
@@ -181,7 +181,7 @@ We may want to compensate for sequencing depth, scaling the transcript abundance
TidyTranscriptomics ```{r normalise, cache=TRUE} -counts_SE.norm = counts_SE.aggr %>% identify_abundant(factor_of_interest = condition) %>% scale_abundance() +counts_SE.norm = counts_SE.aggr |> identify_abundant(factor_of_interest = condition) |> scale_abundance() ```
@@ -201,15 +201,15 @@ norm_counts.table <- cpm(dgList)
```{r, include=FALSE} -counts_SE.norm %>% select(`count`, count_scaled, .abundant, everything()) +counts_SE.norm |> select(`count`, count_scaled, .abundant, everything()) ``` We can easily plot the scaled density to check the scaling outcome. On the x axis we have the log scaled counts, on the y axes we have the density, data is grouped by sample and coloured by cell type. ```{r plot_normalise, cache=TRUE} -counts_SE.norm %>% - ggplot(aes(count_scaled + 1, group=sample, color=`Cell.type`)) + +counts_SE.norm |> + ggplot(aes(count_scaled + 1, group=.sample, color=`Cell.type`)) + geom_density() + scale_x_log10() + my_theme @@ -222,7 +222,7 @@ We may want to identify and filter variable transcripts.
TidyTranscriptomics ```{r filter variable, cache=TRUE} -counts_SE.norm.variable = counts_SE.norm %>% keep_variable() +counts_SE.norm.variable = counts_SE.norm |> keep_variable() ```
@@ -261,7 +261,7 @@ We may want to reduce the dimensions of our data, for example using PCA or MDS a TidyTranscriptomics ```{r mds, cache=TRUE} counts_SE.norm.MDS = - counts_SE.norm %>% + counts_SE.norm |> reduce_dimensions(method="MDS", .dims = 6) ``` @@ -275,7 +275,7 @@ count_m_log = log(count_m + 1) cmds = limma::plotMDS(ndim = .dims, plot = FALSE) cmds = cmds %$% - cmdscale.out %>% + cmdscale.out |> setNames(sprintf("Dim%s", 1:6)) cmds$cell_type = tibble_counts[ @@ -289,10 +289,10 @@ cmds$cell_type = tibble_counts[ On the x and y axes axis we have the reduced dimensions 1 to 3, data is coloured by cell type. ```{r plot_mds, cache=TRUE} -counts_SE.norm.MDS %>% pivot_sample() %>% select(contains("Dim"), everything()) +counts_SE.norm.MDS |> pivot_sample() |> select(contains("Dim"), everything()) -counts_SE.norm.MDS %>% - pivot_sample() %>% +counts_SE.norm.MDS |> + pivot_sample() |> GGally::ggpairs(columns = 6:(6+5), ggplot2::aes(colour=`Cell.type`)) @@ -304,7 +304,7 @@ counts_SE.norm.MDS %>% TidyTranscriptomics ```{r pca, cache=TRUE, message=FALSE, warning=FALSE, results='hide'} counts_SE.norm.PCA = - counts_SE.norm %>% + counts_SE.norm |> reduce_dimensions(method="PCA", .dims = 6) ```
@@ -312,7 +312,7 @@ counts_SE.norm.PCA = Standard procedure (comparative purpose) ```{r,eval=FALSE} count_m_log = log(count_m + 1) -pc = count_m_log %>% prcomp(scale = TRUE) +pc = count_m_log |> prcomp(scale = TRUE) variance = pc$sdev^2 variance = (variance / sum(variance))[1:6] pc$cell_type = counts[ @@ -327,10 +327,10 @@ On the x and y axes axis we have the reduced dimensions 1 to 3, data is coloured ```{r plot_pca, cache=TRUE} -counts_SE.norm.PCA %>% pivot_sample() %>% select(contains("PC"), everything()) +counts_SE.norm.PCA |> pivot_sample() |> select(contains("PC"), everything()) -counts_SE.norm.PCA %>% - pivot_sample() %>% +counts_SE.norm.PCA |> + pivot_sample() |> GGally::ggpairs(columns = 11:13, ggplot2::aes(colour=`Cell.type`)) ``` @@ -339,8 +339,8 @@ counts_SE.norm.PCA %>% TidyTranscriptomics ```{r tsne, cache=TRUE, message=FALSE, warning=FALSE, results='hide'} counts_SE.norm.tSNE = - breast_tcga_mini_SE %>% - identify_abundant() %>% + breast_tcga_mini_SE |> + identify_abundant() |> reduce_dimensions( method = "tSNE", perplexity=10, @@ -371,12 +371,12 @@ tsne$cell_type = tibble_counts[ Plot ```{r} -counts_SE.norm.tSNE %>% - pivot_sample() %>% +counts_SE.norm.tSNE |> + pivot_sample() |> select(contains("tSNE"), everything()) -counts_SE.norm.tSNE %>% - pivot_sample() %>% +counts_SE.norm.tSNE |> + pivot_sample() |> ggplot(aes(x = `tSNE1`, y = `tSNE2`, color=Call)) + geom_point() + my_theme ``` @@ -387,7 +387,7 @@ We may want to rotate the reduced dimensions (or any two numeric columns really) TidyTranscriptomics ```{r rotate, cache=TRUE} counts_SE.norm.MDS.rotated = - counts_SE.norm.MDS %>% + counts_SE.norm.MDS |> rotate_dimensions(`Dim1`, `Dim2`, rotation_degrees = 45, action="get") ```
@@ -399,9 +399,9 @@ rotation = function(m, d) { ((bind_rows( c(`1` = cos(r), `2` = -sin(r)), c(`1` = sin(r), `2` = cos(r)) - ) %>% as_matrix) %*% m) + ) |> as_matrix()) %*% m) } -mds_r = pca %>% rotation(rotation_degrees) +mds_r = pca |> rotation(rotation_degrees) mds_r$cell_type = counts[ match(counts$sample, rownames(mds_r)), "Cell.type" @@ -414,7 +414,7 @@ mds_r$cell_type = counts[ On the x and y axes axis we have the first two reduced dimensions, data is coloured by cell type. ```{r plot_rotate_1, cache=TRUE} -counts_SE.norm.MDS.rotated %>% +counts_SE.norm.MDS.rotated |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`Cell.type` )) + geom_point() + my_theme @@ -424,8 +424,8 @@ counts_SE.norm.MDS.rotated %>% On the x and y axes axis we have the first two reduced dimensions rotated of 45 degrees, data is coloured by cell type. ```{r plot_rotate_2, cache=TRUE} -counts_SE.norm.MDS.rotated %>% - pivot_sample() %>% +counts_SE.norm.MDS.rotated |> + pivot_sample() |> ggplot(aes(x=`Dim1_rotated_45`, y=`Dim2_rotated_45`, color=`Cell.type` )) + geom_point() + my_theme @@ -438,7 +438,7 @@ We may want to test for differential transcription between sample-wise factors o TidyTranscriptomics ```{r de, cache=TRUE, message=FALSE, warning=FALSE, results='hide'} counts_SE.de = - counts_SE %>% + counts_SE |> test_differential_abundance( ~ condition, action="get") counts_SE.de ``` @@ -464,8 +464,8 @@ topTags(qlf, n=Inf) The functon `test_differential_abundance` operated with contrasts too. The constrasts hve the name of the design matrix (generally ) ```{r de contrast, cache=TRUE, message=FALSE, warning=FALSE, results='hide', eval=FALSE} counts_SE.de = - counts_SE %>% - identify_abundant(factor_of_interest = condition) %>% + counts_SE |> + identify_abundant(factor_of_interest = condition) |> test_differential_abundance( ~ 0 + condition, .contrasts = c( "conditionTRUE - conditionFALSE"), @@ -481,7 +481,7 @@ We may want to adjust `counts` for (known) unwanted variation. `adjust_abundance TidyTranscriptomics ```{r adjust, cache=TRUE, message=FALSE, warning=FALSE, results='hide'} counts_SE.norm.adj = - counts_SE.norm %>% adjust_abundance( .factor_unwanted = batch, .factor_of_interest = factor_of_interest) + counts_SE.norm |> adjust_abundance( .factor_unwanted = batch, .factor_of_interest = factor_of_interest) ``` @@ -524,7 +524,7 @@ We may want to infer the cell type composition of our samples (with the algorith TidyTranscriptomics ```{r cibersort, cache=TRUE} counts_SE.cibersort = - counts_SE %>% + counts_SE |> deconvolve_cellularity(action="get", cores=1, prefix = "cibersort__") ``` @@ -534,7 +534,7 @@ Standard procedure (comparative purpose) ```{r, eval=FALSE} source(‘CIBERSORT.R’) -count_m %>% write.table("mixture_file.txt") +count_m |> write.table("mixture_file.txt") results <- CIBERSORT( "sig_matrix_file.txt", "mixture_file.txt", @@ -552,13 +552,13 @@ results$cell_type = tibble_counts[ With the new annotated data frame, we can plot the distributions of cell types across samples, and compare them with the nominal cell type labels to check for the purity of isolation. On the x axis we have the cell types inferred by Cibersort, on the y axis we have the inferred proportions. The data is facetted and coloured by nominal cell types (annotation given by the researcher after FACS sorting). ```{r plot_cibersort, cache=TRUE} -counts_SE.cibersort %>% +counts_SE.cibersort |> pivot_longer( names_to= "Cell_type_inferred", values_to = "proportion", names_prefix ="cibersort__", cols=contains("cibersort__") - ) %>% + ) |> ggplot(aes(x=`Cell_type_inferred`, y=proportion, fill=`Cell.type`)) + geom_boxplot() + facet_wrap(~`Cell.type`) + @@ -572,7 +572,7 @@ We can also perform a statistical test on the differential cell-type abundance a ```{r DC, cache=TRUE} - counts_SE %>% + counts_SE |> test_differential_cellularity(. ~ condition ) ``` @@ -583,16 +583,16 @@ We can also perform regression analysis with censored data (coxph). # Add survival data counts_SE_survival = - counts_SE %>% - nest(data = -sample) %>% + counts_SE |> + nest(data = -sample) |> mutate( days = sample(1:1000, size = n()), dead = sample(c(0,1), size = n(), replace = TRUE) - ) %>% + ) |> unnest(data) # Test -counts_SE_survival %>% +counts_SE_survival |> test_differential_cellularity(survival::Surv(days, dead) ~ .) ``` @@ -602,7 +602,7 @@ We can also perform test of Kaplan-Meier curves. ```{r DC_censored_stratification} counts_stratified = - counts_SE_survival %>% + counts_SE_survival |> # Test test_stratification_cellularity( @@ -629,7 +629,7 @@ We may want to cluster our data (e.g., using k-means sample-wise). `cluster_elem
TidyTranscriptomics ```{r cluster, cache=TRUE} -counts_SE.norm.cluster = counts_SE.norm.MDS %>% +counts_SE.norm.cluster = counts_SE.norm.MDS |> cluster_elements(method="kmeans", centers = 2, action="get" ) ```
@@ -653,7 +653,7 @@ cluster$cell_type = tibble_counts[ We can add cluster annotation to the MDS dimension reduced data set and plot. ```{r plot_cluster, cache=TRUE} - counts_SE.norm.cluster %>% + counts_SE.norm.cluster |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`cluster_kmeans`)) + geom_point() + my_theme @@ -668,7 +668,7 @@ Matrix package (v1.3-3) causes an error with Seurat::FindNeighbors used in this TidyTranscriptomics ```{r SNN, eval=FALSE, cache=TRUE, message=FALSE, warning=FALSE, results='hide'} counts_SE.norm.SNN = - counts_SE.norm.tSNE %>% + counts_SE.norm.tSNE |> cluster_elements(method = "SNN") ``` @@ -699,20 +699,20 @@ snn$cell_type = tibble_counts[
```{r SNN_plot, eval=FALSE, cache=TRUE} -counts_SE.norm.SNN %>% - pivot_sample() %>% +counts_SE.norm.SNN |> + pivot_sample() |> select(contains("tSNE"), everything()) -counts_SE.norm.SNN %>% - pivot_sample() %>% - gather(source, Call, c("cluster_SNN", "Call")) %>% - distinct() %>% +counts_SE.norm.SNN |> + pivot_sample() |> + gather(source, Call, c("cluster_SNN", "Call")) |> + distinct() |> ggplot(aes(x = `tSNE1`, y = `tSNE2`, color=Call)) + geom_point() + facet_grid(~source) + my_theme # Do differential transcription between clusters -counts_SE.norm.SNN %>% - mutate(factor_of_interest = `cluster_SNN` == 3) %>% +counts_SE.norm.SNN |> + mutate(factor_of_interest = `cluster_SNN` == 3) |> test_differential_abundance( ~ factor_of_interest, action="get" @@ -732,7 +732,7 @@ We may want to remove redundant elements from the original data set (e.g., sampl TidyTranscriptomics ```{r drop, cache=TRUE} counts_SE.norm.non_redundant = - counts_SE.norm.MDS %>% + counts_SE.norm.MDS |> remove_redundancy( method = "correlation" ) ``` @@ -750,14 +750,14 @@ library(widyr) sort = TRUE, diag = FALSE, upper = FALSE - ) %>% - filter(correlation > correlation_threshold) %>% - distinct(item1) %>% + ) |> + filter(correlation > correlation_threshold) |> + distinct(item1) |> rename(!!.element := item1) # Return non redudant data frame -counts %>% anti_join(.data.correlated) %>% - spread(sample, rc, - transcript) %>% +counts |> anti_join(.data.correlated) |> + spread(sample, rc, - transcript) |> left_join(annotation) @@ -769,8 +769,8 @@ counts %>% anti_join(.data.correlated) %>% We can visualise how the reduced redundancy with the reduced dimentions look like ```{r plot_drop, cache=TRUE} -counts_SE.norm.non_redundant %>% - pivot_sample() %>% +counts_SE.norm.non_redundant |> + pivot_sample() |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`Cell.type`)) + geom_point() + my_theme @@ -781,7 +781,7 @@ counts_SE.norm.non_redundant %>% ```{r drop2, cache=TRUE} counts_SE.norm.non_redundant = - counts_SE.norm.MDS %>% + counts_SE.norm.MDS |> remove_redundancy( method = "reduced_dimensions", Dim_a_column = `Dim1`, @@ -792,8 +792,8 @@ counts_SE.norm.non_redundant = We can visualise MDS reduced dimensions of the samples with the closest pair removed. ```{r plot_drop2, cache=TRUE} -counts_SE.norm.non_redundant %>% - pivot_sample() %>% +counts_SE.norm.non_redundant |> + pivot_sample() |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`Cell.type`)) + geom_point() + my_theme @@ -824,7 +824,7 @@ counts = tidybulk_SAM_BAM( We can add gene symbols from ensembl identifiers. This is useful since different resources use ensembl IDs while others use gene symbol IDs. This currently works for human and mouse. ```{r ensembl, cache=TRUE} -counts_ensembl %>% ensembl_to_symbol(ens) +counts_ensembl |> ensembl_to_symbol(ens) ``` ## From gene symbol to gene description (gene name in full) @@ -832,8 +832,8 @@ counts_ensembl %>% ensembl_to_symbol(ens) We can add gene full name (and in future description) from symbol identifiers. This currently works for human and mouse. ```{r description} -counts_SE %>% - describe_transcript() %>% +counts_SE |> + describe_transcript() |> select(feature, description, everything()) ``` diff --git a/README.md b/README.md index 278fa030..74c4a5bb 100755 --- a/README.md +++ b/README.md @@ -17,27 +17,24 @@ License. website: [stemangiola.github.io/tidybulk/](http://stemangiola.github.io/tidybulk/) - +[Third party +tutorials](https://rstudio-pubs-static.s3.amazonaws.com/792462_f948e766b15d4ee5be5c860493bda0b3.html) Please have a look also to -- [tidySummarizedExperiment](https://github.com/stemangiola/tidySummarizedExperiment) - for bulk data tidy representation -- [tidySingleCellExperiment](https://github.com/stemangiola/tidySingleCellExperiment) - for single-cell data tidy representation -- [tidyseurat](https://github.com/stemangiola/tidyseurat) for - single-cell data tidy representation -- [tidyHeatmap](https://github.com/stemangiola/tidyHeatmap) for - heatmaps produced with tidy principles analysis and manipulation -- [nanny](https://github.com/stemangiola/nanny) for tidy high-level - data analysis and manipulation -- [tidygate](https://github.com/stemangiola/tidygate) for adding - custom gate information to your tibble +- [tidySummarizedExperiment](https://github.com/stemangiola/tidySummarizedExperiment) + for bulk data tidy representation +- [tidySingleCellExperiment](https://github.com/stemangiola/tidySingleCellExperiment) + for single-cell data tidy representation +- [tidyseurat](https://github.com/stemangiola/tidyseurat) for + single-cell data tidy representation +- [tidyHeatmap](https://github.com/stemangiola/tidyHeatmap) for heatmaps + produced with tidy principles analysis and manipulation +- [tidygate](https://github.com/stemangiola/tidygate) for adding custom + gate information to your tibble + [![Build Status](https://travis-ci.org/stemangiola/tidybulk.svg?branch=master)](https://travis-ci.org/stemangiola/tidybulk) [![Coverage Status](https://coveralls.io/repos/github/stemangiola/tidybulk/badge.svg?branch=master)](https://coveralls.io/github/stemangiola/tidybulk?branch=master) + --> @@ -99,7 +96,7 @@ We will use a `SummarizedExperiment` object counts_SE ``` - ## # A SummarizedExperiment-tibble abstraction: 408,624 × 48 + ## # A SummarizedExperiment-tibble abstraction: 408,624 × 8 ## # Features=8513 | Samples=48 | Assays=count ## .feature .sample count Cell.type time condition batch factor_of_interest ## @@ -113,8 +110,7 @@ counts_SE ## 8 AANAT SRR1740034 284 b_cell 0 d TRUE 0 TRUE ## 9 AAR2 SRR1740034 379 b_cell 0 d TRUE 0 TRUE ## 10 AARS2 SRR1740034 685 b_cell 0 d TRUE 0 TRUE - ## # … with 40 more rows - ## # ℹ Use `print(n = ...)` to see more rows + ## # ℹ 40 more rows Loading `tidySummarizedExperiment` will automatically abstract this object as `tibble`, so we can display it and manipulate it with tidy @@ -135,7 +131,7 @@ First of all, you can cite all articles utilised within your workflow automatically from any tidybulk tibble ``` r -counts_SE %>% get_bibliography() +counts_SE |> get_bibliography() ``` ## Aggregate duplicated `transcripts` @@ -155,7 +151,7 @@ TidyTranscriptomics ``` r rowData(counts_SE)$gene_name = rownames(counts_SE) -counts_SE.aggr = counts_SE %>% aggregate_duplicates(.transcript = gene_name) +counts_SE.aggr = counts_SE |> aggregate_duplicates(.transcript = gene_name) ``` @@ -198,9 +194,11 @@ scaled data as `_scaled`. TidyTranscriptomics ``` r -counts_SE.norm = counts_SE.aggr %>% identify_abundant(factor_of_interest = condition) %>% scale_abundance() +counts_SE.norm = counts_SE.aggr |> identify_abundant(factor_of_interest = condition) |> scale_abundance() ``` + ## tidybulk says: the sample with largest library size SRR1740080 was chosen as reference for scaling +
@@ -229,8 +227,8 @@ the x axis we have the log scaled counts, on the y axes we have the density, data is grouped by sample and coloured by cell type. ``` r -counts_SE.norm %>% - ggplot(aes(count_scaled + 1, group=sample, color=`Cell.type`)) + +counts_SE.norm |> + ggplot(aes(count_scaled + 1, group=.sample, color=`Cell.type`)) + geom_density() + scale_x_log10() + my_theme @@ -247,9 +245,11 @@ We may want to identify and filter variable transcripts. TidyTranscriptomics ``` r -counts_SE.norm.variable = counts_SE.norm %>% keep_variable() +counts_SE.norm.variable = counts_SE.norm |> keep_variable() ``` + ## Getting the 500 most variable genes +
@@ -298,10 +298,14 @@ TidyTranscriptomics ``` r counts_SE.norm.MDS = - counts_SE.norm %>% + counts_SE.norm |> reduce_dimensions(method="MDS", .dims = 6) ``` + ## Getting the 500 most variable genes + + ## tidybulk says: to access the raw results do `attr(..., "internals")$MDS` +
@@ -315,7 +319,7 @@ count_m_log = log(count_m + 1) cmds = limma::plotMDS(ndim = .dims, plot = FALSE) cmds = cmds %$% - cmdscale.out %>% + cmdscale.out |> setNames(sprintf("Dim%s", 1:6)) cmds$cell_type = tibble_counts[ @@ -334,33 +338,42 @@ On the x and y axes axis we have the reduced dimensions 1 to 3, data is coloured by cell type. ``` r -counts_SE.norm.MDS %>% pivot_sample() %>% select(contains("Dim"), everything()) +counts_SE.norm.MDS |> pivot_sample() |> select(contains("Dim"), everything()) ``` ## # A tibble: 48 × 14 - ## Dim1 Dim2 Dim3 Dim4 Dim5 Dim6 .sample Cell.…¹ time condi…² - ## - ## 1 -1.46 0.220 -1.68 0.0553 0.0658 -0.126 SRR17400… b_cell 0 d TRUE - ## 2 -1.46 0.226 -1.71 0.0300 0.0454 -0.137 SRR17400… b_cell 1 d TRUE - ## 3 -1.44 0.193 -1.60 0.0890 0.0503 -0.121 SRR17400… b_cell 3 d TRUE - ## 4 -1.44 0.198 -1.67 0.0891 0.0543 -0.110 SRR17400… b_cell 7 d TRUE - ## 5 0.243 -1.42 0.182 0.00642 -0.503 -0.131 SRR17400… dendri… 0 d FALSE - ## 6 0.191 -1.42 0.195 0.0180 -0.457 -0.130 SRR17400… dendri… 1 d FALSE - ## 7 0.257 -1.42 0.152 0.0130 -0.582 -0.0927 SRR17400… dendri… 3 d FALSE - ## 8 0.162 -1.43 0.189 0.0232 -0.452 -0.109 SRR17400… dendri… 7 d FALSE - ## 9 0.516 -1.47 0.240 -0.251 0.457 -0.119 SRR17400… monocy… 0 d FALSE - ## 10 0.514 -1.41 0.231 -0.219 0.458 -0.131 SRR17400… monocy… 1 d FALSE - ## # … with 38 more rows, 4 more variables: batch , factor_of_interest , - ## # TMM , multiplier , and abbreviated variable names ¹​Cell.type, - ## # ²​condition - ## # ℹ Use `print(n = ...)` to see more rows, and `colnames()` to see all variable names + ## Dim1 Dim2 Dim3 Dim4 Dim5 Dim6 .sample Cell.type time + ## + ## 1 -1.46 0.220 -1.68 0.0553 0.0658 -0.126 SRR1740034 b_cell 0 d + ## 2 -1.46 0.226 -1.71 0.0300 0.0454 -0.137 SRR1740035 b_cell 1 d + ## 3 -1.44 0.193 -1.60 0.0890 0.0503 -0.121 SRR1740036 b_cell 3 d + ## 4 -1.44 0.198 -1.67 0.0891 0.0543 -0.110 SRR1740037 b_cell 7 d + ## 5 0.243 -1.42 0.182 0.00642 -0.503 -0.131 SRR1740038 dendritic_mye… 0 d + ## 6 0.191 -1.42 0.195 0.0180 -0.457 -0.130 SRR1740039 dendritic_mye… 1 d + ## 7 0.257 -1.42 0.152 0.0130 -0.582 -0.0927 SRR1740040 dendritic_mye… 3 d + ## 8 0.162 -1.43 0.189 0.0232 -0.452 -0.109 SRR1740041 dendritic_mye… 7 d + ## 9 0.516 -1.47 0.240 -0.251 0.457 -0.119 SRR1740042 monocyte 0 d + ## 10 0.514 -1.41 0.231 -0.219 0.458 -0.131 SRR1740043 monocyte 1 d + ## # ℹ 38 more rows + ## # ℹ 5 more variables: condition , batch , factor_of_interest , + ## # TMM , multiplier ``` r -counts_SE.norm.MDS %>% - pivot_sample() %>% +counts_SE.norm.MDS |> + pivot_sample() |> GGally::ggpairs(columns = 6:(6+5), ggplot2::aes(colour=`Cell.type`)) ``` + ## Registered S3 method overwritten by 'GGally': + ## method from + ## +.gg ggplot2 + + ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`. + ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`. + ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`. + ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`. + ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`. + ![](man/figures/plot_mds-1.png) **PCA** @@ -371,7 +384,7 @@ TidyTranscriptomics ``` r counts_SE.norm.PCA = - counts_SE.norm %>% + counts_SE.norm |> reduce_dimensions(method="PCA", .dims = 6) ``` @@ -383,7 +396,7 @@ Standard procedure (comparative purpose) ``` r count_m_log = log(count_m + 1) -pc = count_m_log %>% prcomp(scale = TRUE) +pc = count_m_log |> prcomp(scale = TRUE) variance = pc$sdev^2 variance = (variance / sum(variance))[1:6] pc$cell_type = counts[ @@ -402,29 +415,29 @@ On the x and y axes axis we have the reduced dimensions 1 to 3, data is coloured by cell type. ``` r -counts_SE.norm.PCA %>% pivot_sample() %>% select(contains("PC"), everything()) +counts_SE.norm.PCA |> pivot_sample() |> select(contains("PC"), everything()) ``` ## # A tibble: 48 × 14 - ## PC1 PC2 PC3 PC4 PC5 PC6 .sample Cell.…¹ time condi…² batch - ## - ## 1 -12.6 -2.52 -14.9 -0.424 -0.592 -1.22 SRR174… b_cell 0 d TRUE 0 - ## 2 -12.6 -2.57 -15.2 -0.140 -0.388 -1.30 SRR174… b_cell 1 d TRUE 1 - ## 3 -12.6 -2.41 -14.5 -0.714 -0.344 -1.10 SRR174… b_cell 3 d TRUE 1 - ## 4 -12.5 -2.34 -14.9 -0.816 -0.427 -1.00 SRR174… b_cell 7 d TRUE 1 - ## 5 0.189 13.0 1.66 -0.0269 4.64 -1.35 SRR174… dendri… 0 d FALSE 0 - ## 6 -0.293 12.9 1.76 -0.0727 4.21 -1.28 SRR174… dendri… 1 d FALSE 0 - ## 7 0.407 13.0 1.42 -0.0529 5.37 -1.01 SRR174… dendri… 3 d FALSE 1 - ## 8 -0.620 13.0 1.73 -0.201 4.17 -1.07 SRR174… dendri… 7 d FALSE 0 - ## 9 2.56 13.5 2.32 2.03 -4.32 -1.22 SRR174… monocy… 0 d FALSE 1 - ## 10 2.65 13.1 2.21 1.80 -4.29 -1.30 SRR174… monocy… 1 d FALSE 1 - ## # … with 38 more rows, 3 more variables: factor_of_interest , TMM , - ## # multiplier , and abbreviated variable names ¹​Cell.type, ²​condition - ## # ℹ Use `print(n = ...)` to see more rows, and `colnames()` to see all variable names + ## PC1 PC2 PC3 PC4 PC5 PC6 .sample Cell.type time condition + ## + ## 1 -12.6 -2.52 -14.9 -0.424 -0.592 -1.22 SRR17400… b_cell 0 d TRUE + ## 2 -12.6 -2.57 -15.2 -0.140 -0.388 -1.30 SRR17400… b_cell 1 d TRUE + ## 3 -12.6 -2.41 -14.5 -0.714 -0.344 -1.10 SRR17400… b_cell 3 d TRUE + ## 4 -12.5 -2.34 -14.9 -0.816 -0.427 -1.00 SRR17400… b_cell 7 d TRUE + ## 5 0.189 13.0 1.66 -0.0269 4.64 -1.35 SRR17400… dendriti… 0 d FALSE + ## 6 -0.293 12.9 1.76 -0.0727 4.21 -1.28 SRR17400… dendriti… 1 d FALSE + ## 7 0.407 13.0 1.42 -0.0529 5.37 -1.01 SRR17400… dendriti… 3 d FALSE + ## 8 -0.620 13.0 1.73 -0.201 4.17 -1.07 SRR17400… dendriti… 7 d FALSE + ## 9 2.56 13.5 2.32 2.03 -4.32 -1.22 SRR17400… monocyte 0 d FALSE + ## 10 2.65 13.1 2.21 1.80 -4.29 -1.30 SRR17400… monocyte 1 d FALSE + ## # ℹ 38 more rows + ## # ℹ 4 more variables: batch , factor_of_interest , TMM , + ## # multiplier ``` r -counts_SE.norm.PCA %>% - pivot_sample() %>% +counts_SE.norm.PCA |> + pivot_sample() |> GGally::ggpairs(columns = 11:13, ggplot2::aes(colour=`Cell.type`)) ``` @@ -438,8 +451,8 @@ TidyTranscriptomics ``` r counts_SE.norm.tSNE = - breast_tcga_mini_SE %>% - identify_abundant() %>% + breast_tcga_mini_SE |> + identify_abundant() |> reduce_dimensions( method = "tSNE", perplexity=10, @@ -476,30 +489,29 @@ tsne$cell_type = tibble_counts[ Plot ``` r -counts_SE.norm.tSNE %>% - pivot_sample() %>% +counts_SE.norm.tSNE |> + pivot_sample() |> select(contains("tSNE"), everything()) ``` ## # A tibble: 251 × 4 - ## tSNE1 tSNE2 .sample Call - ## - ## 1 -5.25 10.2 TCGA-A1-A0SD-01A-11R-A115-07 LumA - ## 2 6.41 2.79 TCGA-A1-A0SF-01A-11R-A144-07 LumA - ## 3 -9.28 6.63 TCGA-A1-A0SG-01A-11R-A144-07 LumA - ## 4 -1.76 4.82 TCGA-A1-A0SH-01A-11R-A084-07 LumA - ## 5 -1.41 12.2 TCGA-A1-A0SI-01A-11R-A144-07 LumB - ## 6 -1.89 -3.60 TCGA-A1-A0SJ-01A-11R-A084-07 LumA - ## 7 18.5 -13.4 TCGA-A1-A0SK-01A-12R-A084-07 Basal - ## 8 -4.03 -10.4 TCGA-A1-A0SM-01A-11R-A084-07 LumA - ## 9 -2.84 -10.8 TCGA-A1-A0SN-01A-11R-A144-07 LumB - ## 10 -19.3 5.03 TCGA-A1-A0SQ-01A-21R-A144-07 LumA - ## # … with 241 more rows - ## # ℹ Use `print(n = ...)` to see more rows + ## tSNE1 tSNE2 .sample Call + ## + ## 1 -4.29 5.40 TCGA-A1-A0SD-01A-11R-A115-07 LumA + ## 2 4.48 -2.82 TCGA-A1-A0SF-01A-11R-A144-07 LumA + ## 3 -9.06 0.637 TCGA-A1-A0SG-01A-11R-A144-07 LumA + ## 4 7.05 7.28 TCGA-A1-A0SH-01A-11R-A084-07 LumA + ## 5 -2.38 2.77 TCGA-A1-A0SI-01A-11R-A144-07 LumB + ## 6 -1.63 -5.67 TCGA-A1-A0SJ-01A-11R-A084-07 LumA + ## 7 18.2 -18.3 TCGA-A1-A0SK-01A-12R-A084-07 Basal + ## 8 -9.06 -11.7 TCGA-A1-A0SM-01A-11R-A084-07 LumA + ## 9 -8.88 -10.3 TCGA-A1-A0SN-01A-11R-A144-07 LumB + ## 10 -8.30 23.4 TCGA-A1-A0SQ-01A-21R-A144-07 LumA + ## # ℹ 241 more rows ``` r -counts_SE.norm.tSNE %>% - pivot_sample() %>% +counts_SE.norm.tSNE |> + pivot_sample() |> ggplot(aes(x = `tSNE1`, y = `tSNE2`, color=Call)) + geom_point() + my_theme ``` @@ -521,7 +533,7 @@ TidyTranscriptomics ``` r counts_SE.norm.MDS.rotated = - counts_SE.norm.MDS %>% + counts_SE.norm.MDS |> rotate_dimensions(`Dim1`, `Dim2`, rotation_degrees = 45, action="get") ``` @@ -537,9 +549,9 @@ rotation = function(m, d) { ((bind_rows( c(`1` = cos(r), `2` = -sin(r)), c(`1` = sin(r), `2` = cos(r)) - ) %>% as_matrix) %*% m) + ) |> as_matrix()) %*% m) } -mds_r = pca %>% rotation(rotation_degrees) +mds_r = pca |> rotation(rotation_degrees) mds_r$cell_type = counts[ match(counts$sample, rownames(mds_r)), "Cell.type" @@ -556,7 +568,7 @@ mds_r$cell_type = counts[ dimensions, data is coloured by cell type. ``` r -counts_SE.norm.MDS.rotated %>% +counts_SE.norm.MDS.rotated |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`Cell.type` )) + geom_point() + my_theme @@ -568,8 +580,8 @@ counts_SE.norm.MDS.rotated %>% dimensions rotated of 45 degrees, data is coloured by cell type. ``` r -counts_SE.norm.MDS.rotated %>% - pivot_sample() %>% +counts_SE.norm.MDS.rotated |> + pivot_sample() |> ggplot(aes(x=`Dim1_rotated_45`, y=`Dim2_rotated_45`, color=`Cell.type` )) + geom_point() + my_theme @@ -593,7 +605,7 @@ TidyTranscriptomics ``` r counts_SE.de = - counts_SE %>% + counts_SE |> test_differential_abundance( ~ condition, action="get") counts_SE.de ``` @@ -630,8 +642,8 @@ The constrasts hve the name of the design matrix (generally ``` r counts_SE.de = - counts_SE %>% - identify_abundant(factor_of_interest = condition) %>% + counts_SE |> + identify_abundant(factor_of_interest = condition) |> test_differential_abundance( ~ 0 + condition, .contrasts = c( "conditionTRUE - conditionFALSE"), @@ -656,7 +668,7 @@ TidyTranscriptomics ``` r counts_SE.norm.adj = - counts_SE.norm %>% adjust_abundance( ~ factor_of_interest + batch) + counts_SE.norm |> adjust_abundance( .factor_unwanted = batch, .factor_of_interest = factor_of_interest) ```
@@ -710,14 +722,10 @@ TidyTranscriptomics ``` r counts_SE.cibersort = - counts_SE %>% + counts_SE |> deconvolve_cellularity(action="get", cores=1, prefix = "cibersort__") ``` - ## - ## The downloaded binary packages are in - ## /var/folders/zn/m_qvr9zd7tq0wdtsbq255f8xypj_zg/T//RtmpIi5KN6/downloaded_packages -
@@ -726,7 +734,7 @@ Standard procedure (comparative purpose) ``` r source(‘CIBERSORT.R’) -count_m %>% write.table("mixture_file.txt") +count_m |> write.table("mixture_file.txt") results <- CIBERSORT( "sig_matrix_file.txt", "mixture_file.txt", @@ -752,13 +760,13 @@ proportions. The data is facetted and coloured by nominal cell types (annotation given by the researcher after FACS sorting). ``` r -counts_SE.cibersort %>% +counts_SE.cibersort |> pivot_longer( names_to= "Cell_type_inferred", values_to = "proportion", names_prefix ="cibersort__", cols=contains("cibersort__") - ) %>% + ) |> ggplot(aes(x=`Cell_type_inferred`, y=proportion, fill=`Cell.type`)) + geom_boxplot() + facet_wrap(~`Cell.type`) + @@ -766,6 +774,8 @@ counts_SE.cibersort %>% theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5), aspect.ratio=1/5) ``` + ## tidySummarizedExperiment says: A data frame is returned for independent data analysis. + ![](man/figures/plot_cibersort-1.png) ## Test differential cell-type abundance @@ -774,31 +784,27 @@ We can also perform a statistical test on the differential cell-type abundance across conditions ``` r - counts_SE %>% + counts_SE |> test_differential_cellularity(. ~ condition ) ``` - ## - ## The downloaded binary packages are in - ## /var/folders/zn/m_qvr9zd7tq0wdtsbq255f8xypj_zg/T//RtmpIi5KN6/downloaded_packages - ## # A tibble: 22 × 7 - ## .cell_type cell_t…¹ estim…² estim…³ std.e…⁴ stati…⁵ p.valu…⁶ - ## - ## 1 cibersort.B.cells.naive -2.94 2.25 0.367 6.13 8.77e-10 - ## 2 cibersort.B.cells.memory -4.86 1.48 0.436 3.40 6.77e- 4 - ## 3 cibersort.Plasma.cells -5.33 -0.487 0.507 -0.960 3.37e- 1 - ## 4 cibersort.T.cells.CD8 -2.33 0.924 0.475 1.94 5.18e- 2 - ## 5 cibersort.T.cells.CD4.naive -2.83 -0.620 0.531 -1.17 2.43e- 1 - ## 6 cibersort.T.cells.CD4.memo… -2.46 0.190 0.500 0.380 7.04e- 1 - ## 7 cibersort.T.cells.CD4.memo… -3.67 2.23 0.427 5.22 1.80e- 7 - ## 8 cibersort.T.cells.follicul… -5.68 -0.217 0.507 -0.427 6.69e- 1 - ## 9 cibersort.T.cells.regulato… -5.04 1.94 0.360 5.39 6.86e- 8 - ## 10 cibersort.T.cells.gamma.de… -4.78 -0.250 0.514 -0.486 6.27e- 1 - ## # … with 12 more rows, and abbreviated variable names ¹​cell_type_proportions, - ## # ²​`estimate_(Intercept)`, ³​estimate_conditionTRUE, ⁴​std.error_conditionTRUE, - ## # ⁵​statistic_conditionTRUE, ⁶​p.value_conditionTRUE - ## # ℹ Use `print(n = ...)` to see more rows + ## .cell_type cell_type_proportions `estimate_(Intercept)` + ## + ## 1 cibersort.B.cells.naive -2.94 + ## 2 cibersort.B.cells.memory -4.86 + ## 3 cibersort.Plasma.cells -5.33 + ## 4 cibersort.T.cells.CD8 -2.33 + ## 5 cibersort.T.cells.CD4.naive -2.83 + ## 6 cibersort.T.cells.CD4.memory.re… -2.46 + ## 7 cibersort.T.cells.CD4.memory.ac… -3.67 + ## 8 cibersort.T.cells.follicular.he… -5.68 + ## 9 cibersort.T.cells.regulatory..T… -5.04 + ## 10 cibersort.T.cells.gamma.delta -4.78 + ## # ℹ 12 more rows + ## # ℹ 4 more variables: estimate_conditionTRUE , + ## # std.error_conditionTRUE , statistic_conditionTRUE , + ## # p.value_conditionTRUE We can also perform regression analysis with censored data (coxph). @@ -806,41 +812,55 @@ We can also perform regression analysis with censored data (coxph). # Add survival data counts_SE_survival = - counts_SE %>% - nest(data = -sample) %>% + counts_SE |> + nest(data = -sample) |> mutate( days = sample(1:1000, size = n()), dead = sample(c(0,1), size = n(), replace = TRUE) - ) %>% + ) |> unnest(data) +``` + + ## Warning in is_sample_feature_deprecated_used(.data, .cols): + ## tidySummarizedExperiment says: from version 1.3.1, the special columns + ## including sample/feature id (colnames(se), rownames(se)) has changed to + ## ".sample" and ".feature". This dataset is returned with the old-style + ## vocabulary (feature and sample), however we suggest to update your workflow to + ## reflect the new vocabulary (.feature, .sample) + ## Warning in is_sample_feature_deprecated_used(.data, .cols): + ## tidySummarizedExperiment says: from version 1.3.1, the special columns + ## including sample/feature id (colnames(se), rownames(se)) has changed to + ## ".sample" and ".feature". This dataset is returned with the old-style + ## vocabulary (feature and sample), however we suggest to update your workflow to + ## reflect the new vocabulary (.feature, .sample) + +``` r # Test -counts_SE_survival %>% +counts_SE_survival |> test_differential_cellularity(survival::Surv(days, dead) ~ .) ``` ## # A tibble: 22 × 6 - ## .cell_type cell_t…¹ estim…² std.e…³ stati…⁴ p.value - ## - ## 1 cibersort.B.cells.naive -0.224 0.415 -0.540 0.589 - ## 2 cibersort.B.cells.memory 0.510 0.346 1.48 0.140 - ## 3 cibersort.Plasma.cells 0.892 0.449 1.99 0.0467 - ## 4 cibersort.T.cells.CD8 0.531 0.639 0.831 0.406 - ## 5 cibersort.T.cells.CD4.naive 0.112 0.386 0.290 0.772 - ## 6 cibersort.T.cells.CD4.memory.resting 0.498 0.540 0.921 0.357 - ## 7 cibersort.T.cells.CD4.memory.activa… 2.37 0.939 2.52 0.0117 - ## 8 cibersort.T.cells.follicular.helper -0.544 0.421 -1.29 0.197 - ## 9 cibersort.T.cells.regulatory..Tregs. 1.59 0.656 2.42 0.0157 - ## 10 cibersort.T.cells.gamma.delta 0.510 0.688 0.741 0.459 - ## # … with 12 more rows, and abbreviated variable names ¹​cell_type_proportions, - ## # ²​estimate, ³​std.error, ⁴​statistic - ## # ℹ Use `print(n = ...)` to see more rows + ## .cell_type cell_type_proportions estimate std.error statistic p.value + ## + ## 1 cibersort.B.cells… 5.15 1.58 3.27 0.00108 + ## 2 cibersort.B.cells… 2.12 1.48 1.43 0.153 + ## 3 cibersort.Plasma.… 2.96 1.35 2.20 0.0279 + ## 4 cibersort.T.cells… 3.94 1.71 2.30 0.0215 + ## 5 cibersort.T.cells… 3.34 1.75 1.91 0.0560 + ## 6 cibersort.T.cells… -0.785 0.868 -0.904 0.366 + ## 7 cibersort.T.cells… -3.15 1.65 -1.91 0.0568 + ## 8 cibersort.T.cells… -0.435 0.421 -1.03 0.301 + ## 9 cibersort.T.cells… 0.795 0.757 1.05 0.294 + ## 10 cibersort.T.cells… -0.0292 0.641 -0.0456 0.964 + ## # ℹ 12 more rows We can also perform test of Kaplan-Meier curves. ``` r counts_stratified = - counts_SE_survival %>% + counts_SE_survival |> # Test test_stratification_cellularity( @@ -852,21 +872,22 @@ counts_stratified ``` ## # A tibble: 22 × 6 - ## .cell_type cell_t…¹ .low_…² .high…³ pvalue plot - ## - ## 1 cibersort.B.cells.naive 14.4 7.56 0.506 - ## 2 cibersort.B.cells.memory 17.2 4.77 0.500 - ## 3 cibersort.Plasma.cells 13.3 8.73 0.903 - ## 4 cibersort.T.cells.CD8 13.9 8.06 0.369 - ## 5 cibersort.T.cells.CD4.naive 12.8 9.15 0.407 - ## 6 cibersort.T.cells.CD4.memory.rest… 7.65 14.4 0.105 - ## 7 cibersort.T.cells.CD4.memory.acti… 15.7 6.26 0.392 - ## 8 cibersort.T.cells.follicular.help… 17.1 4.88 0.949 - ## 9 cibersort.T.cells.regulatory..Tre… 13.7 8.35 0.771 - ## 10 cibersort.T.cells.gamma.delta 16.2 5.76 0.379 - ## # … with 12 more rows, and abbreviated variable names ¹​cell_type_proportions, - ## # ²​.low_cellularity_expected, ³​.high_cellularity_expected - ## # ℹ Use `print(n = ...)` to see more rows + ## .cell_type cell_type_proportions .low_cellularity_exp…¹ + ## + ## 1 cibersort.B.cells.naive 9.41 + ## 2 cibersort.B.cells.memory 10.5 + ## 3 cibersort.Plasma.cells 12.5 + ## 4 cibersort.T.cells.CD8 11.0 + ## 5 cibersort.T.cells.CD4.naive 8.40 + ## 6 cibersort.T.cells.CD4.memory.re… 9.09 + ## 7 cibersort.T.cells.CD4.memory.ac… 11.2 + ## 8 cibersort.T.cells.follicular.he… 13.7 + ## 9 cibersort.T.cells.regulatory..T… 8.16 + ## 10 cibersort.T.cells.gamma.delta 14.8 + ## # ℹ 12 more rows + ## # ℹ abbreviated name: ¹​.low_cellularity_expected + ## # ℹ 3 more variables: .high_cellularity_expected , pvalue , + ## # plot Plot Kaplan-Meier curves @@ -892,7 +913,7 @@ clustering methods. TidyTranscriptomics ``` r -counts_SE.norm.cluster = counts_SE.norm.MDS %>% +counts_SE.norm.cluster = counts_SE.norm.MDS |> cluster_elements(method="kmeans", centers = 2, action="get" ) ``` @@ -924,7 +945,7 @@ We can add cluster annotation to the MDS dimension reduced data set and plot. ``` r - counts_SE.norm.cluster %>% + counts_SE.norm.cluster |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`cluster_kmeans`)) + geom_point() + my_theme @@ -944,7 +965,7 @@ TidyTranscriptomics ``` r counts_SE.norm.SNN = - counts_SE.norm.tSNE %>% + counts_SE.norm.tSNE |> cluster_elements(method = "SNN") ``` @@ -982,20 +1003,20 @@ snn$cell_type = tibble_counts[
``` r -counts_SE.norm.SNN %>% - pivot_sample() %>% +counts_SE.norm.SNN |> + pivot_sample() |> select(contains("tSNE"), everything()) -counts_SE.norm.SNN %>% - pivot_sample() %>% - gather(source, Call, c("cluster_SNN", "Call")) %>% - distinct() %>% +counts_SE.norm.SNN |> + pivot_sample() |> + gather(source, Call, c("cluster_SNN", "Call")) |> + distinct() |> ggplot(aes(x = `tSNE1`, y = `tSNE2`, color=Call)) + geom_point() + facet_grid(~source) + my_theme # Do differential transcription between clusters -counts_SE.norm.SNN %>% - mutate(factor_of_interest = `cluster_SNN` == 3) %>% +counts_SE.norm.SNN |> + mutate(factor_of_interest = `cluster_SNN` == 3) |> test_differential_abundance( ~ factor_of_interest, action="get" @@ -1012,10 +1033,9 @@ symbols; for `sample`, `transcript` and `count`) and returns a tibble with redundant elements removed (e.g., samples). Two redundancy estimation approaches are supported: -- removal of highly correlated clusters of elements (keeping a - representative) with method=“correlation” -- removal of most proximal element pairs in a reduced dimensional - space. +- removal of highly correlated clusters of elements (keeping a + representative) with method=“correlation” +- removal of most proximal element pairs in a reduced dimensional space. **Approach 1** @@ -1025,13 +1045,11 @@ TidyTranscriptomics ``` r counts_SE.norm.non_redundant = - counts_SE.norm.MDS %>% + counts_SE.norm.MDS |> remove_redundancy( method = "correlation" ) ``` - ## - ## The downloaded binary packages are in - ## /var/folders/zn/m_qvr9zd7tq0wdtsbq255f8xypj_zg/T//RtmpIi5KN6/downloaded_packages + ## Getting the 8513 most variable genes @@ -1051,14 +1069,14 @@ library(widyr) sort = TRUE, diag = FALSE, upper = FALSE - ) %>% - filter(correlation > correlation_threshold) %>% - distinct(item1) %>% + ) |> + filter(correlation > correlation_threshold) |> + distinct(item1) |> rename(!!.element := item1) # Return non redudant data frame -counts %>% anti_join(.data.correlated) %>% - spread(sample, rc, - transcript) %>% +counts |> anti_join(.data.correlated) |> + spread(sample, rc, - transcript) |> left_join(annotation) ``` @@ -1072,8 +1090,8 @@ We can visualise how the reduced redundancy with the reduced dimentions look like ``` r -counts_SE.norm.non_redundant %>% - pivot_sample() %>% +counts_SE.norm.non_redundant |> + pivot_sample() |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`Cell.type`)) + geom_point() + my_theme @@ -1085,7 +1103,7 @@ counts_SE.norm.non_redundant %>% ``` r counts_SE.norm.non_redundant = - counts_SE.norm.MDS %>% + counts_SE.norm.MDS |> remove_redundancy( method = "reduced_dimensions", Dim_a_column = `Dim1`, @@ -1097,8 +1115,8 @@ We can visualise MDS reduced dimensions of the samples with the closest pair removed. ``` r -counts_SE.norm.non_redundant %>% - pivot_sample() %>% +counts_SE.norm.non_redundant |> + pivot_sample() |> ggplot(aes(x=`Dim1`, y=`Dim2`, color=`Cell.type`)) + geom_point() + my_theme @@ -1135,26 +1153,26 @@ different resources use ensembl IDs while others use gene symbol IDs. This currently works for human and mouse. ``` r -counts_ensembl %>% ensembl_to_symbol(ens) +counts_ensembl |> ensembl_to_symbol(ens) ``` ## # A tibble: 119 × 8 - ## ens iso `read count` sample cases…¹ cases…² trans…³ ref_g…⁴ - ## - ## 1 ENSG00000000003 13 144 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 2 ENSG00000000003 13 72 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 3 ENSG00000000003 13 0 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 4 ENSG00000000003 13 1099 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 5 ENSG00000000003 13 11 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 6 ENSG00000000003 13 2 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 7 ENSG00000000003 13 3 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 8 ENSG00000000003 13 2678 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 9 ENSG00000000003 13 751 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## 10 ENSG00000000003 13 1 TARGET-20… Acute … Primar… TSPAN6 hg38 - ## # … with 109 more rows, and abbreviated variable names - ## # ¹​cases_0_project_disease_type, ²​cases_0_samples_0_sample_type, ³​transcript, - ## # ⁴​ref_genome - ## # ℹ Use `print(n = ...)` to see more rows + ## ens iso `read count` sample cases_0_project_dise…¹ cases_0_samples_0_sa…² + ## + ## 1 ENSG… 13 144 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 2 ENSG… 13 72 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 3 ENSG… 13 0 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 4 ENSG… 13 1099 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 5 ENSG… 13 11 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 6 ENSG… 13 2 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 7 ENSG… 13 3 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 8 ENSG… 13 2678 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 9 ENSG… 13 751 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## 10 ENSG… 13 1 TARGE… Acute Myeloid Leukemia Primary Blood Derived… + ## # ℹ 109 more rows + ## # ℹ abbreviated names: ¹​cases_0_project_disease_type, + ## # ²​cases_0_samples_0_sample_type + ## # ℹ 2 more variables: transcript , ref_genome ## From gene symbol to gene description (gene name in full) @@ -1162,25 +1180,31 @@ We can add gene full name (and in future description) from symbol identifiers. This currently works for human and mouse. ``` r -counts_SE %>% - describe_transcript() %>% +counts_SE |> + describe_transcript() |> select(feature, description, everything()) ``` - ## # A SummarizedExperiment-tibble abstraction: 408,624 × 48 + ## Warning in is_sample_feature_deprecated_used(.data, .cols): + ## tidySummarizedExperiment says: from version 1.3.1, the special columns + ## including sample/feature id (colnames(se), rownames(se)) has changed to + ## ".sample" and ".feature". This dataset is returned with the old-style + ## vocabulary (feature and sample), however we suggest to update your workflow to + ## reflect the new vocabulary (.feature, .sample) + + ## # A SummarizedExperiment-tibble abstraction: 408,624 × 10 ## # Features=8513 | Samples=48 | Assays=count - ## feature sample count Cell.…¹ time condi…² batch facto…³ descr…⁴ gene_…⁵ - ## - ## 1 A1BG SRR1740034 153 b_cell 0 d TRUE 0 TRUE alpha-… A1BG - ## 2 A1BG-AS1 SRR1740034 83 b_cell 0 d TRUE 0 TRUE A1BG a… A1BG-A… - ## 3 AAAS SRR1740034 868 b_cell 0 d TRUE 0 TRUE aladin… AAAS - ## 4 AACS SRR1740034 222 b_cell 0 d TRUE 0 TRUE acetoa… AACS - ## 5 AAGAB SRR1740034 590 b_cell 0 d TRUE 0 TRUE alpha … AAGAB - ## 6 AAMDC SRR1740034 48 b_cell 0 d TRUE 0 TRUE adipog… AAMDC - ## 7 AAMP SRR1740034 1257 b_cell 0 d TRUE 0 TRUE angio … AAMP - ## 8 AANAT SRR1740034 284 b_cell 0 d TRUE 0 TRUE aralky… AANAT - ## 9 AAR2 SRR1740034 379 b_cell 0 d TRUE 0 TRUE AAR2 s… AAR2 - ## 10 AARS2 SRR1740034 685 b_cell 0 d TRUE 0 TRUE alanyl… AARS2 - ## # … with 40 more rows, and abbreviated variable names ¹​Cell.type, ²​condition, - ## # ³​factor_of_interest, ⁴​description, ⁵​gene_name - ## # ℹ Use `print(n = ...)` to see more rows + ## feature sample count Cell.type time condition batch factor_of_interest + ## + ## 1 A1BG SRR1740034 153 b_cell 0 d TRUE 0 TRUE + ## 2 A1BG-AS1 SRR1740034 83 b_cell 0 d TRUE 0 TRUE + ## 3 AAAS SRR1740034 868 b_cell 0 d TRUE 0 TRUE + ## 4 AACS SRR1740034 222 b_cell 0 d TRUE 0 TRUE + ## 5 AAGAB SRR1740034 590 b_cell 0 d TRUE 0 TRUE + ## 6 AAMDC SRR1740034 48 b_cell 0 d TRUE 0 TRUE + ## 7 AAMP SRR1740034 1257 b_cell 0 d TRUE 0 TRUE + ## 8 AANAT SRR1740034 284 b_cell 0 d TRUE 0 TRUE + ## 9 AAR2 SRR1740034 379 b_cell 0 d TRUE 0 TRUE + ## 10 AARS2 SRR1740034 685 b_cell 0 d TRUE 0 TRUE + ## # ℹ 40 more rows + ## # ℹ 2 more variables: description , gene_name diff --git a/man/figures/plot_cibersort-1.png b/man/figures/plot_cibersort-1.png index ed786285..88be62f5 100644 Binary files a/man/figures/plot_cibersort-1.png and b/man/figures/plot_cibersort-1.png differ diff --git a/man/figures/plot_cluster-1.png b/man/figures/plot_cluster-1.png index bbaeaaed..50773f4b 100644 Binary files a/man/figures/plot_cluster-1.png and b/man/figures/plot_cluster-1.png differ diff --git a/man/figures/plot_drop-1.png b/man/figures/plot_drop-1.png index fdaaf789..67e73cdb 100644 Binary files a/man/figures/plot_drop-1.png and b/man/figures/plot_drop-1.png differ diff --git a/man/figures/plot_drop2-1.png b/man/figures/plot_drop2-1.png index 12f943fc..b2bea833 100644 Binary files a/man/figures/plot_drop2-1.png and b/man/figures/plot_drop2-1.png differ diff --git a/man/figures/plot_mds-1.png b/man/figures/plot_mds-1.png index 3be8cef6..78c69a32 100644 Binary files a/man/figures/plot_mds-1.png and b/man/figures/plot_mds-1.png differ diff --git a/man/figures/plot_normalise-1.png b/man/figures/plot_normalise-1.png index 19092d3e..7a882512 100644 Binary files a/man/figures/plot_normalise-1.png and b/man/figures/plot_normalise-1.png differ diff --git a/man/figures/plot_pca-1.png b/man/figures/plot_pca-1.png index 208dbfec..a5b66ab9 100644 Binary files a/man/figures/plot_pca-1.png and b/man/figures/plot_pca-1.png differ diff --git a/man/figures/plot_rotate_1-1.png b/man/figures/plot_rotate_1-1.png index 3d50c395..348b056f 100644 Binary files a/man/figures/plot_rotate_1-1.png and b/man/figures/plot_rotate_1-1.png differ diff --git a/man/figures/plot_rotate_2-1.png b/man/figures/plot_rotate_2-1.png index a9aba09d..012019dd 100644 Binary files a/man/figures/plot_rotate_2-1.png and b/man/figures/plot_rotate_2-1.png differ diff --git a/man/figures/unnamed-chunk-14-1.png b/man/figures/unnamed-chunk-14-1.png index bfa30c18..ae313d7f 100644 Binary files a/man/figures/unnamed-chunk-14-1.png and b/man/figures/unnamed-chunk-14-1.png differ diff --git a/man/figures/unnamed-chunk-19-1.png b/man/figures/unnamed-chunk-19-1.png index 24447bbe..e6072ba1 100644 Binary files a/man/figures/unnamed-chunk-19-1.png and b/man/figures/unnamed-chunk-19-1.png differ