whtns / chevreul Goto Github PK

[Yesterday 11:43 PM] Shayler, Dominic
Hardeep's 2015 dataset doesn't have the Fetal_Age of his samples in it by default. I've pulled them from another one of my older merged datasets to merge into my new sets. However, every time I try to load the CSV the app crashes. Am I doing it wrong? I thought you had to select the shared column name you wanted to merge and load the .csv?

http://10.134.4.105:3838/production-apps/monocle_heatmaps_SHL_1st_exp/ Is showing an error after you put different shCtrl PTs.

Can I get a full list of genes & transcripts list that are differentially expressed in branch B compared to A? As you generated genes detected before, obviously there are more genes detected in B, so I'd like to do some GO analysis with those genes, even though most of them are super lowly expressed.

          > They have changed the structure of the Seurat object with sc@assays$RNA@layers$counts rather than sc@assays$RNA@counts, so the save isn't recognising this difference when saving.

Is it possible to have a version with the 'layers' bit added to the save and load Seurat Object?

It is possible to "patch" each of SaveH5Seurat, LoadH5Seurat functions (tested with Seurat v5.0.1).

Here it is for SaveH5Seurat:

#' Patch of SeuratDisk::SaveH5Seurat function
#'
#' The "Assay5" class attribute "RNA" needs to be converted to a standard "Assay"
#' class for compatibility with SeuratDisk. It requires to make a temporary copy
#' so the size of the object grows bigger.
#'
#' @param object the Seurat object
#' @param filename the file path where to save the Seurat object
#' @param verbose SaveH5Seurat verbosity
#' @param overwrite whether to overwrite an existing file
#' 
#' @return NULL
SaveH5SeuratObject <- function(
    object,
    filename,
    verbose = TRUE,
    overwrite = TRUE
    ) {
  
  # add copy of "RNA" 
  object[["RNA-tmp"]] <- CreateAssayObject(counts = object[["RNA"]]$counts)
  # remove original
  object[["RNA"]] <- NULL
  # export
  SaveH5Seurat(object, filename = filename, overwrite, verbose)
  
  return(NULL)
}

And for LoadH5Seurat:

#' Patch of SeuratDisk::LoadH5Seurat function
#'
#' The "Assay" class attribute "RNA" needs to be converted to the new "Assay5"
#' class. It requires to make a temporary copy so the size of the object grows
#' bigger.
#'
#' @param filename the file path where to save the Seurat object
#' @param verbose LoadH5Seurat verbosity
#' 
#' @return NULL
LoadH5SeuratObject <- function(
    filename,
    verbose = TRUE
) {
  
  # load h5 data
  object <- LoadH5Seurat(filename)
  # create "Assay5" class from old "Assay" class
  keys <- Key(object)
  slotID <- names(keys)[startsWith(keys, "rna")]
  object[["RNA"]] <- CreateAssay5Object(counts = object[[slotID]]$counts)
  # delete "Assay" class
  object[[slotID]] <- NULL
  # reorder assays list
  object@assays <- object@assays[sort(names(object@assays))]

  return(object)
}

The biggest caveat is that it requires to make a copy of the RNA assay at each pass (saving or loading). A direct conversion would be preferable, but I have not found such a function in the SeuratObject package.

EDIT: A better function would iterate over all assays and do what I did for RNA for each "Assay5" class assay.

Originally posted by @Gilquin in mojaveazure/seurat-disk#172 (comment)

​[12:37 AM] Shayler, Dominic
One other issue came up: I'm trying to merge all the fetal projects with the fetal/organoid 2018 project to get a fetal organoid combined plot. This integration crashes each time so I'm concerned I may have hit the memory limit because I've added a sizable chunk of extra cells. Would you be able to run this?

• Document snakemake workflow;

• link to cobriniklab repo

• make flowchart with ggflowchart or any other tool.

• should list:

  • input file type (.fastq, .bam, etc.)
  • name of tool (trimgalore, etc.)
  • description of tool (trim adapter sequences and low quality bases from sequencing reads)
  • relevant custom parameters
  • output file type (.bam, .gtf, etc.) 5. next step of output file.

• expected output: fastqc > trimgalore > HISAT2 > stringtie > tximport > seurat; additional: megadepth, scanpy, velocyto

https://satijalab.org/seurat/v3.0/visualization_vignette.html

update to selectize and pull list of genes from seurat object

When I run, I get the following error:

Error in convert_symbols_by_species(src_genes = rownames(seu$gene), src_species = "human") :
object 'human_to_mouse_homologs' not found

* fix any broken bullet points
* remove 'site' header
* make home page and getting started the same document

* link to working interactive version of the app
* review the utility of create_project_db
* fix seurat demo link that doesn't work

• review included data
• add description of steps to view analysis details
• move entire site under cobriniklab github user instead of whtns
  * add warning about usage and pitfalls

mutate(coalesce_var = coalesce(!!! syms(vars_select(names(.), starts_with("my")))))

• better checks for which organism is used for a given seurat object (solution in bioconductor?)
  outline:

1. integrate within species (human or mouse)
2. integrate across species (human and mouse)
3. the steps of integration here
4. assumptions of the input data

• enough cells of each batch (>50?)
• the same feature type (symbols, ensgenes, entrezids, etc.)

5. limitations

• data can't be too big

stretch: improve species checking in integration_workflow; refer to annotation types

Hi!

I'm trying to integrate mouse and human data together and I came across this project! Looks really awesome.

One thing I wanted to check was, if I could import my existing seurat objects directly into this tool? A few of these datasets have been around for a long time with all the compute done, so it would be really difficult to reanalyze them.

Regards,
Harish

Dearest Kevin,
We are molecular biology PhD students trying to use the seuratTools package in R (4.1.1 )
with Platform: x86_64-w64-mingw32/x64 (64-bit)
Nevertheless we struggle to install the package and the message error is as such;
Your current architecture is 32bit; however, this version of Python was compiled for 64bit.
Use reticulate::install_miniconda() if you'd like to install a Miniconda Python environment.
ERROR: loading failed

We were wandering if we may receive some guidance and troubleshooting

Sincerely,
Derek

* change x-axis on transcript composition plot
* move TLDR section to 'Getting Started'

• clean up prose
• mention which dataset we are using up top
• change variable names in assorted functions from 'metavar' or 'color.by' to 'group_by'
• format all function names with backticks
• run spellcheck
• make plots smaller?

Transcript numbers detected in branch A,B and C <- Just like you generated genes detected (figure attached!).

1. convert_seu_to_cds

• need to switch b/w RNA and integrated datasets
• need to do feature selection on non-integrated dataset
• smartly deal with paritions (to get multiple pseudotimes; more modules)

1. gene level plots
2. subsetting
3. plotly integration

Errors

1. Monocle doesn't run on organoid app

I recently installed the seuratTools package as an attempt to streamline using Seurat and ComplexHeatmap together. However when I attempt to run the seu_complex_heatmap command I am met with the following error:

Error in as_mapper(.f, ...) : object 'set_names' not found
4.as_mapper(.f, ...)
3.purrr::map2(., ha_col_names.factor, set_names)
2.purrr::map(ha_col_names.factor, ~(scales::hue_pal())(length(.x))) %>% purrr::map2(ha_col_names.factor, set_names)
1.seu_complex_heatmap(oxy_rat_na, features = genes, group.by = "sample_condition")

Any advice as to how to fix this issue would be very appreciated. For reference, my data is a standard Seurat object and I am specifying a list of genes that the heatmap is meant to visualize. The commands I am using are as follows:

genes <- c("List_of_Genes")
seu_complex_heatmap(seurat_object, features=genes, group.by="sample_condition")

Where "sample_condition" is the metadata slot I want my columns to be grouped by. Thank you in advance for whatever help can be offered.

Would it be possible to get dendrogram from Louvain cluster? As far as I understand, Louvain is based on Graph clustering representing non-linear structure, so I assume that it might not be possible, but I'd like to ask you just in case.

We shouldn't rely on projects that mitali or I have created unless we are certain about them. I'd like to move any project with >1 batch to the integrated projects directory to reduce duplication and transfer responsibility to the individual researcher.

​[Yesterday 11:43 PM] Shayler, Dominic
Another question: The RNA Velocity doesn't work correct? Is there any way you might be able to run it on a set for me?

change selectizeinput to 'cluster' for all louvain cluster selections:

• dimredplot
• coverage plot
• heatmap
• violinplot
• etc.

some sort of error during conversion of suerat to monocle

Loading a new project or dataset from a given project shouldn't require restart of the app

Why can't it be installed？

devtools::install_github("whtns/seuratTools")
Downloading GitHub repo whtns/seuratTools@HEAD
whtns-seuratTools-0504522/data/panc8.rda: truncated gzip input
tar.exe: Error exit delayed from previous errors.
Error: Failed to install 'seuratTools' from GitHub:
Does not appear to be an R package (no DESCRIPTION)

One more question about GSEA. After I uploaded csv file and run the GSEA, the server's disconnected. Could you also have a look? I used 20200312-SHL-singlecell-Hs dataset. Thanks

Wondering if I can use "Find Marker" function with branch info instead of Louvain clusters?

I got
Error: 'install_package' is not an exported object from 'namespace:devtools'
when I tried to install this package in R. Any one can help?

1. All transcripts not visible (only 2/5 for NRL in 4seq and 6seq)
2. scale bar not pegged at min 0 in 4seq NRL

1. just show one cluster plot
2. reformat TLDR
3. change all data to human_gene_transcript_seu
4. reformat tables
5. put minimalApp in TLDR; link to working app
6. split code across lines to limit scroll bars in output chunks
7. clean up the last few chunks