search

dviraran / SingleR

SingleR: Single-cell RNA-seq cell types Recognition (legacy version)
GNU General Public License v3.0
263 stars 98 forks source link

Installation of SingleR version 0.2.1. And number of cells using Seurat vs SingleR package #106

Open KeshavB02 opened 5 years ago

KeshavB02 commented 5 years ago

Hi SingleR Team, Can you help in these two problems ? -

  1. -> I want to install SingleR package's 0.2.1 version, becasue for an old project I used 0.2.1 version. I couldn't find any option on the site to install older versions of SingleR. Could you please provide command for that ?

  2. -> Regarding number of cells for 10X data, Using Seurat(2.3.4) I am getting one more cell than SingleR(0.2.2) object, so i am not able to map both the data properly. I have used same parameter for minimum number of genes. Can you please help me in this ? For your reference I am writing the commands for both the packages -

    Seurat (a) pbmc = CreateSeuratObject(raw.data = pbmc.data, min.cells = 3, project = "Sample") (b) pbmc <- FilterCells(object = pbmc, subset.names = "nGene", low.thresholds = 500, high.thresholds = 7500)

    **SingleR ** 
(a) singler = CreateSinglerObject(counts=pbmc@data, annot = NULL, 
project.name=project_name, min.genes = 499, technology = "10X", species = "Human", 
citation = "",  normalize.gene.length = F, variable.genes = "de",  fine.tune = T, do.signatures 
= F, clusters = NULL, do.main.types = T,   reduce.file.size = T ,ref.list = list(c(hpca)))

I have tried 500,501 option for genes in singler & seurat earlier. I think, Seurat package usage <500 & SingleR <=499. So I have written code like that.

Thank-you

dviraran commented 5 years ago

I didn't really maintain versions in a way you can roll back. Sorry. I am not sure now why you have this difference. Not sure what you are using this for, but anyhow, you can just extract the labels/scores from your old object. Why do you need to rerun it with the old version?

KeshavB02 commented 5 years ago

In earlier version, I was getting same cells from both Seurat(2.3.4) & SingleR(0.2.1). So I made full pipeline (for research purpose) according that. Now It is giving error, because of one cell difference. That's why, I want to know, if older version of Singler will work for my data or not.