Dear developer,

I have problem when running NormalyzerDE as values below 1 leads to warnings that this will cause negative values during processing. But my data is not log transform, so I don't want to run using the "noLogTransform=TRUE" option assigned - I want NormalyzerDE to log-transform my data.

The reason values below 1 leads to issues in NormalyzerDE is that they after log transformation will become negative.

If these values are rare you could filter out these entries by doing the following. I am using the package readr for loading and writing the updated data - it could also be done with read.table but then you need to carefully look through what options to use.

> library(readr)  # Load the library readr I use to easily read and write tsv-files
> df <- read_tsv("dataset.tsv")  # Load the dataset
> rows_above_one <- apply(df[, -1], 1, min) > 1  # Calculate which rows have only values above one, the df[,-1] is assuming that your first column contains annotation - if more you need to trim them too
> df_above_one <- df[rows_above_one, ]  # Select the rows that passed the threshold
> write_tsv(df_above_one, path="dataset_above_one.tsv")  # Write a new file

This new file can now be used as input to NormalyzerDE.

library(NormalyzerDE)
normalyzer(jobName="my_job", designPath="design.tsv", dataPath="dataset_above_one.tsv")