know about level sample compare result

[aman21392]

I have direct-RNA sample nanopore data one is control and another is treated sample. So I just want to know the m6A modification in my treated sample. I used human_transcript.fasta file for resquiggle command -- "tombo resquiggle single_fast5/ homotranscript.fa --overwrite --processes 50" Then I used tombo detect modifications level_sample_compare --fast5-basedirs treated/single_fast5/ --alternate-fast5-basedirs control/single_fast5/ --minimum-test-reads 1 --processes 50 --num-most-significant-stored 14000000 --statistics-file-basename sample.level_samp_comp_detect

Then I used command-- tombo text_output browser_files --fast5-basedirs treated/single_fast5/ --control-fast5-basedirs control/single_fast5/ --statistics-filename sample.level_samp_comp_detect.tombo.stats --browser-file-basename sample.level_samp_comp_detect --file-types statistic --motif-descriptions DRACH:3:dam_6mA I got 2 wiggle file. plus and minus and I think only plus wig file we mainly use. I show you my result of plus.wig file looks like--- variableStep chrom=ENST00000203630 span=1 531 0.8333 532 1.0000 533 1.0000 534 1.0000 535 1.0000 536 0.8333 537 0.8333 538 0.8333 539 0.8333 540 0.8333 541 0.8333 542 1.0000 543 1.0000 544 1.0000 545 1.0000 546 1.0000 547 1.0000 548 0.8333 549 0.8333 550 0.8333 551 0.8333 552 0.8333 553 0.6667 554 0.6667 555 0.5000 556 0.5000 557 0.6667 558 0.8333 559 1.0000 560 1.0000 561 1.0000 562 1.0000 563 1.0000 564 1.0000 565 1.0000 566 0.8333 567 0.6667 568 0.5000 569 0.6667 570 0.8333 571 1.0000 572 0.8333 573 0.8333 574 0.8333 575 0.8333 576 0.8333 577 0.6667 578 0.8333 579 0.8333 580 0.8333 581 0.6667 582 0.6667 583 0.8333

I run all command correctly or not .Please just me help to understand which region is m6A modification what is it means of wiggle output and what these column means. Thanks in advance

[marcus1487]

Yes all commands appear correct. The tombo detect_modifications level_sample_compare defaults to the KS-test and stored the effect size statistic by default (for the KS test this is the D-statistic). The larger the value of this statistic the greater likelihood of a modified base in general. See more information about this in the docs here.

I would warn away from setting --minimum-test-reads 1. The statistics produced by the sample comparison methods generally require higher coverage in order to be useful (hence the default of 50X minimum coverage). Lower coverage is likely the reason for the large number of positives (likely false positives) in the presented snippet of your results.

As for specific detection of 6mA this is left to the user to determine the identity of the modified base. Tombo provides tools and a framework to identify consistent shifts in raw nanopore signal. The interpretation of these shifts is left to the user.

[aman21392]

Thanks for you reply. I using --minimum-test-reads 1 because i think if I have some low coverage of reads than we don't want to miss those reads. I uses RNA sample and i just think may I have some RNA which have lower reads. So i don't want to loose them. If I used --minimum-test-reads 1 than I don't get perfectly modification result. What you say. Thanks in advance

[FerchoHQ]

Hi, I have a question regarding to "tombo detect_modifications model_sample_compare", what means --control-fast5-basedirs? how do you get this.

Thanks for your attention.