Incorrect Accuracy Calculation in GTN Tutorial

[Achukwunta]

Hello GTN Team,

While co-instructing the Genomics and Bioinformatics course for 4th-year students, I came across a possible mistake in one of your tutorials titled “Quality Control.” Since the majority of the students had no coding experience, we used Galaxy tutorials as a beginner-friendly approach to sequencing data analysis. As I was explaining Phred quality scores and accuracy calculations, I noticed that the tutorial calculates accuracy using the formula Accuracy = 100—Probability error (P). However, the correct formula should be Accuracy = 1—Probability error (P) x 100.

Another approach is to convert the probability error to % error and then subtract the value from 100. In that same solution, the answer to question number 4 (the corresponding nucleotide G has an accuracy of almost 99.96%) is incorrect.

Why This Matters:

1. It raises the accuracy values, leading to overestimated accuracy in sequencing quality, which can be misleading.
2. The standard approach in bioinformatics and sequencing quality assessment follows (1–P) x100 rather than 100–P.

This is the link to the Galaxy tutorial:
https://training.galaxyproject.org/training-material/topics/sequence-analysis/tutorials/quality-control/tutorial.html

I appreciate all the effort GTN puts to make the platform and the tutorials open-source, free, and accessible to all.

Thanks,
Augustine

[bgruening]

Thanks @Achukwunta for the detailed issue description. We will look at this next week!