gangwug / metacycle Goto Github PK

The input format description is severely lacking. Can you post some sample input files (e.g. sample CSV files) and the corresponsing filestyle parameters used to load them?

Hi Dr. Wu,

I used metacycle to analyze a RNA-seq dataset, which has 4 integer time points and 3 replicates in each time point.

The command line I used is

meta2d(infile="allmerged_colnames.csv", filestyle = "csv",outdir="JTK",timepoints="Line1",cycMethod="JTK")

The metacycle output generates two P value for each transcript - adj P and BH-Q. To my understanding, adj P was calculated by Bonferroni test and BH-q was calculated by Benjamini-Hochberg.

I am wondering which P value would you recommend to use? I am asking because my BH-Q test only give either 1 or 0.59 for all transcripts, which means probably none of these genes are cycling, but adj P can give me about 1000 cycling transcript if I set the cutoff to < 0.05.

Thank you!

Yang

hello，

I don't known why

Hi, thanks for providing this wonderful tool, but I am wandering (1) can this be applied to RNA-seq data? (2) should the expression counts be normalized in the 2d expression matrix? for example, in the form of raw counts, or log transformed data, or RPKM, etc.? (3) can this tool be applied to large scale data sets such as single cell RNAseq, say, 2000 cells for each time point, and 6 time points in total? greatly appreciate your reply!

my code like this:
runARS(fpkm,ARStime = 24,minper = 20,maxper = 28,arsper = 24,arsmet = "mle",releaseNote = T)
my data of fpkm like this:
head(fpkm )
GeneID 2 6 10 14 18 22
1 ENSMUSG00000000001 39.2423000 53.30970000 45.74855000 48.16060000 50.86520000 43.88400000
2 ENSMUSG00000000028 0.7999640 1.04892201 1.03209000 0.88276200 0.75456200 0.60465828
3 ENSMUSG00000000037 0.0647347 0.04107064 0.07675234 0.02379874 0.04734312 0.06447354
4 ENSMUSG00000000049 5.3027700 9.22718000 7.01244000 5.88129000 6.76613000 7.76443900
5 ENSMUSG00000000056 9.3720300 11.69575600 10.73397000 9.05826000 9.78742000 7.47848200
6 ENSMUSG00000000058 21.7220940 21.10826826 23.26603702 25.86059600 26.37826686 23.11869876

with error message like this:

 Error in if (set_order == length(self.x)) { : 
  missing value where TRUE/FALSE needed 
3.
get_period(per.x = eva.x, per.dt_y = eva.dt_y, per.delta = eva.delta, 
    per.pID = eva.pID, is_filter = p1, ar_method = p2) 
2.
evaluate(eva.x = time_points, eva.y = dataM[line, ], eva.delta = self.delta, 
    eva.pID = line, T_start = start, T_end = end, T_default = arsper, 
    arsmethods = arsmet) 
1.
runARS(fpkm, ARStime = 24, minper = 20, maxper = 28, arsper = 24, 
    arsmet = "mle", releaseNote = T) 

best wish.

Hi, Dr. Gang Wu,
Thank you for sharing such powful tool. My data is 0.5-hour-resolution with missing values, I ran JTK_cycle on my data.

from the docs, JTK is not suitable for non-integer interval.
There is one line in the Rscript to set interval length: jtk.init(period, interval), I set the interval to be 1, and also 0.5, other parameters in the script are same, it turned out that I got the same rhythmic genes.
My question is: if the interval is non-integer, maybe I can treat it as an interval(like one hour), the interval length should not rhythmic pattern?

Thank you very much for your help in advance!

Best,
Yichen Hu

Hello Dr. Wu,

I've tried running Meta2D two different ways:

meta2d(infile="NormCounts.Introns.Quant.txt", filestyle="txt",
outdir="IntronQuants", timepoints="line1",
cycMethod=c("JTK","LS"))

which returns the error:

Error in meta2d(infile = "NormCounts.Introns.Quant.txt", filestyle = "txt", :
Please replace non-numeric character or 'NA' with numeric value (time points) in the first line of 'infile'.

and I've also tried it:

meta2d(infile="NormCounts.Introns.Quant.txt", filestyle="txt",
outdir="IntronQuants", timepoints=rep(seq(2, 22, by=4), each=2)
cycMethod=c("JTK","LS"))

which returns the error:

Error in meta2d(infile = "NormCounts.Introns.Quant.txt", filestyle = "txt", :
Please check 'infile' or 'timepoints', the column number of 'infile' should be one more than the length of time points. See the example data in this package

My line 1 looks like this:

geneID 2 2 6 6 10 10 14 14 18 18 22 22

I imagine it's something very basic I'm doing wrong, but I can't figure out for the life of me what it is. I've tried csv and txt files, I've tried directly copy and pasting the first line from the sample files into my input files, but nothing seems to work. Would you happen to have any ideas?

Hi!

Thank you for this package, it seems really exciting.
I am trying to run it (1day, 6 time points, 3 replicates).
When I run

outD <- meta2d(infile="~/Analysis/RNAseq/data_out/counts.normalized.csv", outdir = "~/Analysis/RNAseq/data_out/", filestyle="csv", 
               timepoints="line1", 
               minper = 8,
               maxper = 24, 
               cycMethod = c("ARS", "JTK", "LS"),
               analysisStrategy = "auto",
               outputFile = TRUE,
               outIntegration = "both",
               adjustPhase = "predictedPer",
               combinePvalue = "fisher", 
               weightedPerPha = FALSE,
               ARSmle = "auto",
               ARSdefaultPer = 24, 
               outRawData = FALSE, 
               releaseNote = TRUE,
               outSymbol = "", 
               parallelize = FALSE, 
               inDF = NULL)

I get this error message:

Error in meta2d(infile = "~/Analysis/RNAseq/data_out/counts.normalized.csv",  : 
  lazy-load database '/Users/p/Library/R/3.6/library/MetaCycle/R/MetaCycle.rdb' is corrupt
In addition: Warning messages:
1: In meta2d(infile = "~/Analysis/RNAseq/data_out/counts.normalized.csv",  :
  restarting interrupted promise evaluation
2: In meta2d(infile = "~/Analysis/RNAseq/data_out/counts.normalized.csv",  :
  internal error -3 in R_decompress1

I formatted the first line of the csv file (row.names=F,col.names=T,quote=F,sep=",") like this:
ProbeID,1,1,1,5,5,5,9,9,9,13,13,13,17,17,17,21,21,21

Any tips?

Thanks!

Peter

why does the result of meta2d function show nan value?

JTK_period JTK_adjphase JTK_amplitude meta2d_Base meta2d_AMP meta2d_rAMP
0 NA 0 0 NA NA

Hi Dr. Wu,

while I was running meta_2d function, an error raised: Error in -JTK.ALT[[alt.id]]$EXV : invalid argument to unary operator
my data contains missing values, sample intervals are 30 min, intested period is 1080 min, so my codes are
meta2d(infile = 'foo.csv', outdir = 'foo', filestyle = 'csv', timepoints = 'line1', minper = 1020, maxper = 1140, cycMethod = c("ARS", "JTK", "LS"), analysisStrategy = "auto", ARSdefaultPer = 1080)

I know my data is not suitable for ARS method, I just set the default parameters.
Details:

R version 4.0.0
MetaCycle_1.2.0
foo.txt
I hava also attached input file here. It was comma seperated, could you please kindly help me to check what was going wrong, thank you very much!

Best,
Hu Yichen

Hello,

I have been running meta2d with the following command:

meta2d(infile="my file.csv", timepoints="line1", filestyle="csv", maxper=26, cycMethod=c("JTK","LS"))

I have used this command on 6 different files that are all formatted identically with an identical gene list. The timepoints are in the top row, and the gene names are in the first column.

When I run this command on these files, 5 of 6 files run perfectly. On the 6th file, I get the following error:

Any ideas what might be causing this error and how I could go about fixing it?

Thanks for any advice you might have!

Dear all,

First of all, many thanks to the author. Great work.

I have been wondering if there is any known reason why not to implement the meta2d (or even the meta3d) functions to receive data.frame (even better if melted).

I have been playing with the code and I can already input data.frames in the format especified for the text files. Would it be interesting for the community to have the meta2d altered to receive melted tables?

Best wishes,
Thiago