Hello! I’ve created a metro map for Bacass. Could you give me some feedback? Is it easy to follow? Is there anything you would improve? I still need to add the version tag, so it will be 2.6.0 (next release)

Generally looks great I think! Some questions/suggestions: • I think 3) Genome assembly could be 2) because its visually before Read Taxonomic classification? • Is the Kraken2 output indeed used downstream? I thought thats not used anywhere? Ah now: its displayed in multiqc?! Hm, now it makes the impression that its used in an earlier process somehow. • Kmerfinder is similar, I would rather let it merge into QUAST, because thats where it can be used? • Medaka is shown as default but the blue line also skips that, what should that mean? I would expect its going through Medaka in any way (but that can be opted-out) • in 5) are 4 options but just 3 dots? • Whats the difference between "MultiQC" and "MultiQC custom" in 6)? Is that relevant?

Thanks @Daniel Straub, Yes, I agree with the ordering between genome assembly and tax classification. It will be visually more consistent

Now I'm checking if kraken2 is now showing in the multiqc step. Overall, yes... I agree it can be misleading. Would it make more sense to connect kraken2 directly to multiqc?

Yes, if that visually possible, I think that would be clearer

Kmerfinder is currently connected to “QUAST by refseq”.

Do you mean simplifying this into a single QUAST step instead of separating “QUAST”, “QUAST by refseq”, etc.?

Ah right, my bad. No, all good with kmerfinder then I think.

The current representation could bea bit misleading. Would it be better to if I adjust it so Medaka is clearly on the main path and skipping is shown as an opt-out.?

Yes, I think Medaka should be on the main path and a dot with inner dot (i.e. default).

Well, it’s not really that relevant.

The I would suggest to keep just one entry of those two.

Got it, I’ll work on it 🙌

more nit-picky points; • the dots in the default nodes are not centered correctly • i would align the tools with their nodes (if possible, for example in the genome assembly section) and dropt the

/

in all of them • the fast5 input looks a bit off, try to have it to the left of minimap, so you don't need to go vertical • make the distance between the blue lines for medaka and minimap equidistant. • don't hide the yello line behind the other lines, show them all in equal width • instead of ending by going to the left, try going to the right to make the data flow easier to grasp • you misspelled parallel in the legend 🙂 • the input and output symbol in the legend have different proportions then in the acutal map

Hello!! @Daniel Straub @mashehu Thanks for the feedback. I have added your suggestions and I did my best to align all the dots/nodes and edges 😅 .

sorry, still not centered enough 🙂 what tool are you using?

app.diagrams.net , which has buttons to align objects but I think it is not accomplishing at all.

more nit-picks: • legend and groups have different background color • imo the fast5 input can stay in the 3 grouping, avoid the long and winding line. • I would try to put all group labels above the nodes (5 will be trickiest, but should be possible) • shouldn't minimap2+nanopolish be two nodes? • in one the labels for the nodes on the blue line are too far away from them • is unicycler always run in parallel with the other assembly tools? because it is used in all, you might be able to move it to the front, seperate from the others • overall in a very nice shape already!

Hello!! Sorry for the long wait. Here’s an updated version with improvements to the metro map. (@mashehu)

shouldn't minimap2+nanopolish be two node?

My bad, I made a mistake here. Minimap is already part of the miniasm step, an nanopolish would be an additional node for the long reads (it also relies on minimap) when FAST5 files are available. Since minimap is required in both miniasm and nanopolish steps, i will omit it as a separate node

is unicycler always run in parallel with the other assembly tools? because it is used in all, you might be able to move it to the front, seperate from the others

You are right, I have tried to fix it by combining lines where they share common modules.