[ffavero]

I binned the chromosome 22, but still get this error:

> library('sequenza')

> data.file <- "binned_22.seqz.gz"
> seqz.data <- read.seqz(data.file, chr.name = "22")
> test <- sequenza.extract(data.file)
Processing 22: 1293 variant calls; 36633 heterozygous positions;
2311953 homozygous positions.

> cp22.example=sequenza.fit(test)
  |                                                                      |   0%
Error in result$L - max.lik : non-numeric argument to binary operator
In addition: There were 50 or more warnings (use warnings() to see 
the first 50)

The file is at: http://caprica.thruhere.net:8080/binned_22.seqz.gz

It was created with the following command:
sequenza-utils.py seqz-binning -w 200 -a 22.seqz | /home/ltfang/apps/tabix/bgzip > binned_22.seqz.gz



Chromosome 1 ran fine. However, some commands did not work:
...

I fixed various bugs on the sequenza-utils.py code, and I re-analysed the cell-lines from TCGA benckmark4.
the fixed code is here:

After binning and merging the chromosomes, I've used sequenza, the R package, with this parameters:

sqz     <- "seqz/HCC1143-100_bin200.seqz.gz"
dir.out <- "out"
sqz.id  <- "HCC1143-100"
library(sequenza)

test <- sequenza.extract(file = sqz, gamma = 60, min.reads = 20,
                                  kmin = 200, weighted.mean = FALSE,
                                  min.fw.freq = 0, min.reads.baf = 15)
CP <- sequenza.fit(sequenza.extract = test, chromosome.list = 1:23,
                          XY = c(X = "X", Y = "Y"), ratio.priority = FALSE,
                          female = TRUE, mc.cores=4)

sequenza.results(sequenza.extract = test, cp.table = CP, 
                        ratio.priority = FALSE, XY = c(X = "X", Y = "Y"),
                        out.dir = file.path(dir.out, sqz.id),
                        sample.id = sqz.id, female = TRUE)

I have cellularity 100% and ploidy 3.6 for HCC1143, and cellularity 100% and ploidy 4.5 for HCC1954. CNV match the SKY karyotype.

I also analyzed the mixture and it worked as well. As far as cellularity bigger than 30% we can detect ploidy correctly.

[lethalfang]

I see. Good point.
I'll give it another try in the next week or two, when I have time.