Using BMiner with BTyper3 outputs

[domenico-simone]

Hi,

I am interested in using BMiner with BTyper3 results, but when I upload the BTyper3 final results on the RStudio interface all the tabs in the Shinyapp report no data for any gene/feature. However, I was able to upload BTyper (v2.3.4) final results. Are BTyper3 results compatible with BMiner?

Thank you,

Domenico

[lmc297]

Hi Domenico,

BTyper3 results sadly aren't compatible with BMiner (I'll update the manual to explicitly say this), which only works with the original BTyper (versions 2.X.X); I changed the output format of files produced by BTyper3 to a tabular format based on feedback from users, with the hope that it would be easier for people to work with.

You can get a table of all of your BTyper3 results which you can then load into R, Excel, etc., simply by:

1. Moving to your output directory: cd < name of the directory you supplied to btyper3 -o >
2. Concatenating the BTyper3 final results files into a single file, here named btyper3_all_results.txt (the -v option excludes the file headers): cat btyper3_final_results/*_final_results.txt | grep -v "#filename" > btyper3_all_results.txt

If you want something similar to the MLST/panC/species bar plots which BMiner produces, you should be able to load your table into R, view it in Excel, etc. to make plots.

If you want the virulence factor presence/absence PCA/NMDS plots which BMiner produces, I would recommend running your genomes through the original BTyper (versions 2.X.X) and loading the results into BMiner (the only differences between BTyper and BTyper3 virulence factor detection are the default BLAST thresholds used).

My apologies for the confusion! I'll make a note to provide users with some R/Python scripts to re-create BMiner plots with BTyper3 results, but it will likely be quite some time before I will have time to do this.

[domenico-simone]

Hi Laura,

thank you for your reply and your hints!

I was thinking to contribute to the code and craft a script that could create a input for BMiner from BTyper3 outputs. I have only a question: does BTyper3 provide AMR genes and rpoB genes predictions? I couldn't find them in the output folder.

Thank you,

Domenico

[lmc297]

Hi Domenico,
I opted to exclude the rpoB allelic typing and AMR gene detection methods when I developed BTyper3; rpoB was excluded because we found that the resolution rpoB provided was lower than that of MLST and panC (I might add it back in a future version if users request it, but for now, I opted to leave it out), and the AMR determinant detection method was excluded solely because it was difficult to maintain, wasn't B. cereus group-specific, and because there are so many AMR determinant detection methods out there now, including some that do effectively the same thing as the original BTyper (version 2.X.X) (thanks for your message...I'll definitely also add this to the manual!)