I'm analyzing pseudobulked single-cell RNA-seq data using DESeq2 and comparing **apeglm** vs **ashr** for log fold change shrinkage. Even in my largest cell type (comparing 8 vs 11 samples, 25-7771 cells per sample with median...apeglm** appears much more aggressive in shrinkage, pushing many genes toward logFC = 0
- **ashr** maintains a gradient of shrinkage values and preserves more moderate e…
Order by: Date
30 treatments vs mock using DESeq2 (v.1.44.0). I don't filter any low counts. I apply L2FC-shrinkage (ashr).
This is an experiment where we repeated the library prep from exactly the same RNA which was used for a first prep/experiment
updated 14 months ago • fjrsa
dds, contrast = c("conditions", "wild_type_114_2_no_carbon", "wild_type_114_2_glucose"), type = "ashr", alpha = 0.05)
summary(res)
Then I get:
> summary(res)
out of 9010 with nonzero total read count
adjusted p-value < 0.1
LFC >...alpha 0.05 and not the default?
```r
res <- lfcShrink(dds, res=res_wt_glu_vs_nocar, type = "ashr")
```
I thought the ashr method allowed t…
Additionally, is it possible to use apeglm shrinkage for this comparison? or am I limited to ashr?
Any help is much Appreciated. Thanks
updated 19 months ago • BioNovice247
at several other posts here but wasn't able to fix the issue (while keeping the apeglm instead of ashr.
Here is my entire workflow:
```r
mydata<-read.table("data/july9_salmon.merged.gene_counts_length_scaled.tsv",sep...res = dd1) : type='apeglm' shrinkage only for use with 'coef'
```
I need to use the apeglm, not ashr with contrast.
How do I fix this? Any help is highly appreci…
updated 19 months ago • Mark
I have questions about the use of numeric contrast and lfcshrink.
I set up the dds object using:
dds <- DESeqDataSetFromMatrix(countData = counts, colData = metadata, ~ Sex + genotype * treatment)
Model matrix was calculated using:
mod_mat <- model.matrix(design(dds, data = colData(dds))
Model matrix first line and last line show:
![enter image description here][1]
I …
updated 21 months ago • hx
dds)
> res.cort<- lfcShrink(dds, contrast=c("group","Cortex_Rumitin","Cortex_DMSO"), type="ashr")
> res.cort
log2 fold change (MMSE): group Cortex_Rumitin vs Cortex_DMSO
Wald test p-value: group Cortex_Rumitin vs Cortex_DMSO...gt; res.stri<- lfcShrink(dds, contrast=c("group","Striatum_Rumitin","Striatum_DMSO"),type="ashr")
> res.stri
log2 fold change …
updated 21 months ago • shaunpeterson
dds)
resultsNames(dds)
A_WT10vsWT0 <- lfcShrink(dds, contrast=c("group", "WTT10", "WTT0"), type="ashr")
#Make volcano plot plotting FC vs padj
EnhancedVolcano(df_A,
lab = NA,
x = 'log2FoldChange',
y = 'padj',
title = "WT_T10 vs WT_T0",
pCutoff
updated 22 months ago • james.zhang20
For A_vs_C, I followed:
```r
res_A_vs_C= lfcShrink(dds, contrast=c("Conditions", "A", "C"), type='ashr')
```
Now, I want to compare samples with A and B conditions together versus condition C (A+B_vs_C). How can I do that from the above
region1_condition1)
res <- lfcShrink(dds, contrast = contrast, res =res, type="ashr")
```
As a result, I will have four sets of Differentially Expressed Genes (DEGs). Subsequently, I aim to compare each set of DEGs
updated 2.0 years ago • paria.alipour
in advance for your help and insights,
Theo
Edit 1:
The MA plot looks similar with `lfcShrink(type='ashr')`:
![MA plot after LFC shrink, type=ashr][6]
I suspect something is off with the data but I have not been able to track it down...
```
> sessionInfo
updated 2.1 years ago • theophile
method for the comparison of two conditions.
However if i prefer the lfcshrink FC (apeglm or ashr methods) then what would be the value to be used for GSEA? Indeed the "stat" column is generated using the normal method and
updated 2.2 years ago • delphine.rossille
Hi everyone,
I am trying to carry out some analysis for an experiment which has three dimensions (treatment, genotype and time) and while I have managed to extract the pairwise comparisons I am interested in, I would like to go further and look at the genes which behave differently upon my drug treatment depending on either the genotype or the time factor (for a given genotype) - I believe that …
updated 2.6 years ago • james.zhang20
res = res,
type = "ashr")
}
```
Any help or advice, would be really appreciated!
Thanks
updated 2.6 years ago • Abir.khazaal
Hello,
We have RNA-seq as well as functional data of various kinds.
For some of the most important genes in our study, we quantified protein and transcript levels using wet lab assays to confirm the size and direction of effect of the results.
Our original data were from microarray studies, these data showed massive downregulation of interferon responsive genes in the microarray data; su…
Hello,
@rory I have a question regarding the 3 scenarios in which lfcShrink is applied in DiffBind package. The source code looks like this
```
if(shrink) {
if(contrast$contrastType=="byname") {
res$de <- suppressMessages(DESeq2::lfcShrink(pv$DESeq2$DEdata,
coef=contrast$contrast,
…
updated 2.6 years ago • bhandary.8590
the default algorithm that is used is "aplegm", while the lfcShrink function in DiffBind uses ashr. Is this right? and if so, what is the reason for this? It seems that aplegm is more suited for RNA-seq data while ashr is for other...sequencing data. Could you offer more insight as to why ashr is used by DiffBind?
3. For some of the pairwise DiffBind results, we found some very interesting resul…
updated 2.7 years ago • bhandary.8590
Hi everyone,
I am trying to analyse some RNA-Seq data in DESeq2, but am struggling a little conceptually in understanding how exactly to investigate what genotype specific changes I have.
My data consists of 2 genotypes, 3 timepoints and 2 treatments (treated/untreated) - 48 samples in total (4 replicates per). Therefore, in the sample information file, I have formatted it like so:
Sample Geno…
updated 2.7 years ago • james.zhang20
list(c("status_als_vs_con")))
res1 <- lfcShrink(dds, contrast = contrast, res =res1, type="ashr")
''
Thanks,
Paria
updated 2.8 years ago • paria.alipour
changes as well as the visulal representation.
I tested this mostly using apeglm, but when I tried ashr, the absolute values didn't match up either.
There are not too many DGEs in this data set, otherwise I might have missed it
Since I'm analyzing my SLAM-seq data for the first time, I thought I'd follow this guide: https://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#quick-start
In the guide, it says to use contrast (or name) because contrast sets the log fold change to 0. I have two conditions treatment and control. And for my dds$condition, I releveled it so that the ref = "control". …
not redone?
I hope that the question is clear. Also, I already have implemented a for loop with ashr that produces a list (res as in the example above) with all given comparisons.
[1]: https://support.bioconductor.org/p/9147589
updated 3.1 years ago • theodore.georgomanolis
0 expression, the results have much lower number of DEGs.
Also, after updating some packages (as ashr and mixsqp), I am getting lower number of DEGs when running exactly the same code.
Any explanation please?
Many thanks
this sound reasonable?
2. Here is the MA plot from the same dataset used for volcano plots. with ashr shrinkage (I need to use contrast, so I select this method for shrinkage) I wonder if this MA plots looks okay. They are slightly
I got this message every time:
**Error: package or namespace load failed for ‘DESeq2’ in loadNamespace(j <- i[[1L]], c(lib.loc, .libPaths()), versionCheck = vI[[j]]):
there is no package called ‘Biostrings’
In addition: Warning message:
package ‘DESeq2’ was built under R version 4.1.1**
```r
> sessionInfo()
R version 4.1.0 (2021-05-18)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Ru…
listValues = c(1/3,-1)))
resLFC <- lfcShrink(dds, contrast=contrast, res=res, type="ashr
I have been working with DESeq2 for the past couple of months analyzing my data, I have read over the vignette many times, found other workshops, read message boards, but I still second guess my decisions and the options I have chosen.
Basically my design is that I have multiple clam lines, lets say 3 (A,B,C), and two salinities I am comparing (35 ppt vs 15 ppt). Salinity 35 ppt is my control…
updated 3.5 years ago • lgspeight
We have RNAseq data with 4 cell lines with different knockout genotypes (WT, A, G, GA), and for each cell lines, we have two treatment types: treated (H) and untreated (C) with 5 replicates for each condition. we are interested in the treatment effect on each of these cell lines: WT.H_C, A.H_C, G.H_C, or GA.H_C (H_C means treatment vs untreat comparison). These are easily to be done in DESeq2 bas…
updated 3.5 years ago • Mike
Enter the body of text here
Hi, there:
I read the previous posts about deriving confidence interval for fold change using DESeq result at URLs: https://support.bioconductor.org/p/9142141/ and https://support.bioconductor.org/p/80725/#80729
one comment from Micheal Love mentioned:
Estimated standard errors for the estimated coefficients on the log2 scale are given by the lfcSE column. Ye…
updated 3.5 years ago • Mike
c("ConditionB","ConditionC")),
listValues = c(1,-1/2), type="ashr")
using 'ashr' for LFC shrinkage. If used in published research, please cite:
Stephens, M. (2016) False discovery rates: a new deal...the `contrast` or `listValues` or is there likely something wrong with my input data? I reinstalled `ashr` but it doesn't help
updated 3.5 years ago • Kent
My dataset has a total of 53 samples (with outliers removed) of breast cancer patients with two conditions.
I have read that DESeq2 doesn’t provide automatic shrinkage, and I have to use lfcShrink to shrink the log2FoldChanges. The manual states that apeGLM and ashr perform better than the normal option and that apeGLM is quite strict compared to the others and performs well with small...and I h…
updated 3.6 years ago • Riley
control"),listValues = c(1/2,-1)))
resLFC <- lfcShrink(dds, contrast=contrast, res=res, type="ashr")
up <- res[which(resLFC$log2FoldChange > 0 & resLFC$padj < 0.01),]
down <- res[which(resLFC$log2FoldChange < 0 & resLFC
Hi I have done a DESeq2 DGE analysis and it showed an error when I was using ashr shrinkage
```
using 'ashr' for LFC shrinkage. If used in published research, please cite:
Stephens, M. (2016) False discovery rates...contrast = c(-1/12,-1/12,-1/12,-1/12,1,-1/12,-1/12,-1/12,-1/12,-1/12,-1/12,-1/12,-1/12,0), type = "ashr")
```
Edit: Thank you Michael for pointing towards the solution here: htt…
Hi everyone,
I'm currently getting some warnings when using lfcshrinkage with "ashr", regarding solve().
Example
Code should be placed in three backticks as shown below
```
library(zebrafishRNASeq)
data(zfGenes...c("group", "Trt","Ctl"))
ashr_res<-lfcShrink(dds, contrast = c("group", "Trt","Ctl"),type="ashr")
```
This leads to the production of the corresponding output:…
based on Zhu *et al.*, 2018 but am curious as to what criteria others use to justify `apeglm` vs `ashr` for this sort of analysis.
My current dataset is 160 samples, but I'm narrowing focus to 4M + 4F per experimental condition...I have an FDR < 0.05 for ~ 200 genes, but `apeglm` shrinks all my fold changes towards zero, and `ashr` gives LFCs between 1 and 28 for all 200 genes. Base mean…
which is why I don't want to run the model for each coefficient separately.
I also had a look at ashr, but testing against a threshold is not implemented there.
Best,
Frederik
Edit":
```r
ddsShrink_t <- lfcShrink(ddsObj2, contras=c("cell_group","DC3_TLR","DC2_TLR"), type="ashr")
```
It now worked with ashr. But is this correct
equivalent to:
```
res1 <- lfcShrink(dds, contrast=c("condition","treated1","control1"), type="ashr")
res2 <- lfcShrink(dds, contrast=c("condition","treated2","control2"), type="ashr")
```
Given the statistical benefit on the DESeq2
design = ~ sex+month+method+cat)
I've used the default wald test and am applying ashr for LFCshrink which I think is the correct method, however I have previously used IHW for multiple testing correction...of using both please? Or what is used for multiple testing correction otherwise when running with ashr?
Many thanks,
B
updated 4.2 years ago • Bex
3 UP 843
> res_tableOE_shrunk <- lfcShrink(dds, contrast = contrast_oe, type = "ashr")
using 'ashr' for LFC shrinkage. If used in published research, please cite:
Stephens, M. (2016) False discovery rates: a new deal
RcppArmadillo
Suggests: testthat, knitr, rmarkdown, vsn, pheatmap, RColorBrewer,
apeglm, ashr, tximport, tximeta, tximportData, readr, pbapply,
airway, pasilla (>= 0.2.10), glmGamPoi, BiocManager
License: LGPL (>= 3)
MD5sum
updated 4.4 years ago • ladypurrsia
Hi all,
I am using DESeq2 v1.32 for RNAseq analysis. I am using the following code to apply a threshold/ FC cutoff of 1.25 in the wald stat. However, I am seeing some significant genes in res_sig have |LFC| < 1.25. From an old discussion thread in this forum, it seems to be normal if lfcShrink type = "normal". Is it okay in ashr also? Or am I supposed to add a post hoc threshold also? Wil…
updated 4.5 years ago • lamia11me
Hi, Dr love.
I post a question about [weird MAplot or volcano plot of DESeq2 diff result](https://support.bioconductor.org/p/9139060/) and also in [biostar](https://www.biostars.org/p/9484986/#9485059). @atpoint give a useful answer about too many 0 count genes and prefiltering.
It seems that too many 0 count genes makes lfc shrink have a probelm. And I find the `apeglm` and `ashr` result i…
type")
dds <- DESeq(dds)
res_lfc <- lfcShrink(dds = dds,
type = "ashr",
coef = "type_Fx600_vs_Fx593")
plotMA(res_lfc)
as_tibble(res_lfc) %>%
mutate(padj = case_when(
is.na(padj) ~ 1,
TRUE ~ padj
)) %>%
ggplot
updated 4.6 years ago • Guandong Shang
instead of p-adjusted values in results. Is there a way I can get the nominal P-values by using the ashr shrinkage estimator? I only see the s-values.
If not, which P-values is it base on? How can I examine my test by P-value histogram
updated 4.6 years ago • haasroni
Hi, I'm using Deseq2 (v1.26.0) and ashr (v2.2-47) to shrink the log2FoldChange of the genes resulting from interaction term. DEseq2 returns the different shrunken...among Mut.T/UT vs WT.T/UT using the following
**design = ~cell+trt+cell:trt** followed by ashr fold-change shrinkage.
I get different results with ashr with reference levels compared to the ones without defining...for example**).…
I obtained negative svalues from the lfcShrink function of DESeq2 using the ashr shrinkage estimator.
The values are very small and close to 0: the smallest is `-1.110223e-16`, the largest negative is `-1.199592e...I obtained negative svalues from the lfcShrink function of DESeq2 using the ashr shrinkage estimator.
The values are very small and close to 0: the smallest is `-1.110223e-16`, t…
Risso et al. 2014) to estimate
> variables that capture the unwanted variation. In addition, the ashr
> developers have a specific method for accounting for unwanted
> variation in combination with ashr (Gerard and Stephens
updated 5.3 years ago • igor
for LFC shrinkage, the Normal prior from Love et al (2014).
Note that type='apeglm' and type='ashr' have shown to have less bias than type='normal'.
See ?lfcShrink for more details on shrinkage type, and the DESeq2 vignette
updated 5.3 years ago • helenhuang.math
me 1618.
This only happens if I use lfcShrink with "apeglm". It doesn't happen with "normal" or "ashr" or if I just use "results".
Attached is an example of my code and also my ColData for dds1. Due to sequencing issues, I have some
updated 5.3 years ago • serdarT
did comparisons within each culture condition by grouping condition and genotype and using lfcshrink/ashr. For instance for genotype A:
dds <- DESeqDataSetFromMatrix(countData = countdata,
colData = coldata,
design = ~ Experiment...2D contrast
A_2D_v_WT_2D_lfcShrink <- lfcShrink(dds, contrast=c("group"…
updated 5.4 years ago • Nick F
Dear all,
I analyzed RNA-seq data using DESeq2 extracting the results with or without shrinkage
without shrinkage:
res21 <- results(dds2, contrast=c("combined", "XX.DMSO", "XX.dTAG"),
cooksCutoff=FALSE, alpha=0.05)
with shrinkage:
resLFC1 <- lfcShrink(dds2, contrast=c("combined", "XX.DMSO", "XX.dTAG"),
alp…
ashr" generally performs better than "normal" and that it preserves large LFC, but these LFC are unrealistically large with...ashr. I have 2 questions:
1. Is this expected behavior or does log2FC on this scale imply there is something fundamentally wrong...with my design or samples?
2. If this is expected behavior, should I go with "ashr" or "normal"? I am not using apeglm because I need the …
I have created 12 pairs of DGE for my 3x2 design matrix with deseq2
and 10/12 shrinking by ashr works fine
however,for 2 of the pairs, this message appears
> Optimization failed to converge. Results may be unreliable...increasing maxiter and rerunning.
I saw error in a thread here
https://github.com/stephens999/ashr/issues/76
Does installing mixsqp to resolve this issue?
Am I m…
updated 5.8 years ago • kavator
Hi,
I have pretty simple but probably weird question. I have complex design with interaction terms in my model (see below), I run DESeq2_1.20.0 on rna-seq data. It seems that I can't change my contrasts to end up using only apeglm shrinkage and lfcThreshold, instead I'm stuck with using ashr shrinkage method without being able to incorporate lfcThreshold into calculation of null.
What bothe…
updated 5.8 years ago • Iryna
for LFC shrinkage, the Normal prior from Love et al (2014).
Note that type='apeglm' and type='ashr' have shown to have less bias than type='normal'.
See ?lfcShrink for more details on shrinkage type, and the DESeq2 vignette
updated 5.9 years ago • junli1988
the column “stat” , which seems to be lost after the lfcshrink calculation using either apeglm or ashr. My question therefore is:
Q1: How can I still apply the fdrtool in a correct way? Or am I doing something wrong? I do not think
updated 5.9 years ago • Paul
I have RNA seq data an experiment consisting of an untreated and treated samples in triplicates; a total of 6 samples. The culturing of the used tissue can be challenging so a batch effect is expected. I would like to apply apeglm shrinkage and use adjusted p-values as a cutoff to explain effects between conditions, but am encountering some problems. My questions are mainly related to the lfc sh…
for LFC shrinkage, the Normal prior from Love et al (2014).`
Note that type='apeglm' and type='ashr' have shown to have less bias than type='normal'. See ?lfcShrink for more details on shrinkage type, and the DESeq2 vignette
updated 6.1 years ago • nabiyogesh
1]
```
> resLFC_sh2 <- lfcShrink(dds, contrast=c("Treatment", "Sh2", "Control"), type="ashr")
using 'ashr' for LFC shrinkage. If used in published research, please cite:
Stephens, M. (2016) False discovery rates: a new deal
updated 6.5 years ago • MatthewP
Recent ...
Replies
Answer: Inconsistencies between accessor functions for TxDb databases with selected seql
by
a2a2hd42aa
•
0
Technical inconsistencies in database accessors can disrupt your workflow, but finding a reliable source to download online mobile games sh…
Comment: Can sample-level weights be used when analysing GSVA scores in limma?
by
A.Barden
▴
20
Thanks both for this useful information.
Comment: Reproducibility of code in various Bioconductor versions
by
Gordon Smyth
53k
SNPlocs.Hsapiens.dbSNP155.GRCh38 is available for download from the same location as always. It has not changed since Bioc 3.20:
Presumabl…
Comment: Changing multiple display parameters of a specific exon in GeneRegionTrack
by
a2a2hd42aa
•
0
Similarly, for structural complexity in the [He777 Game][1] APK, having precise control over every detail is what makes a high-performance …
Answer: Issue loading and downloading Bioconductor 3.22 packages after update
by
strawberry3
•
0
Hi, thank you all for the advice.
Using James's advice, I managed to install SNPlocs.Hsapiens.dbSNP155.GRCh38, but it took almost half an …
Votes
Awards
• All
Popular Question to
Gordon Smyth
53k
Popular Question to
Matan G.
▴
60
Popular Question to
Jaqueline
•
0
Popular Question to
Dario Strbenac
★
1.6k
Popular Question to
chirag.parsania
▴
10
Locations
• All
The city by the bay, 11 minutes ago
WEHI, Melbourne, Australia, 2 hours ago
United States, 2 hours ago
Austria, 4 hours ago
Germany/Heidelberg, 5 hours ago
The city by the bay, 11 minutes ago
WEHI, Melbourne, Australia, 2 hours ago
United States, 2 hours ago
Austria, 4 hours ago
Germany/Heidelberg, 5 hours ago
Traffic: 935 users visited in the last hour
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by the
version 2.3.6
version 2.3.6