Hello

[hugheb]

I tried to JBrowse2 to align two plant genomes (1 Gb). I got the following error
Error: Cannot create a string longer than 0x1fffffe8 characters

[cmdcolin]

Hi there,
How did you create the alignment? And how large was the resulting output file? Is it PAF format?

JBrowse 2 does have to load the entire file into memory, and large alignments can exceed browser limits (1fffffe8->~532 megabytes)

Even if they can be loaded, it may be slow. I would recommend possibly filtering out small alignments.

You can also consider using a MCScan type alignment (just the protein sequences) instead of a whole genome alignment, as the datasets for MCScan are much smaller and much less tripped up by repetitive sequences (https://github.com/tanghaibao/jcvi/wiki/MCscan-(Python-version))

Down the line, we are working on some things to make loading large whole genome alignments better by using indexing. If you are interested in trying that out as a beta branch I can try to help describe more. That work is here #3859

[hugheb]

Dear Colin I installed JBrowse2 on my computer (MacBook pro, 2.3 GHz 18-Core Intel Xeon W; 256 GB 2666 MHz DDR4) I uploaded the reference genome as a fasta file (964.2 Mb) and a draft genome of another accession of the same species (fasta file size is 1Gb). I used the Type as IndexedFastaAdapter which is default. I want to find out what scaffolds from the draft genome match chromosome 7 of my reference genome. would it make a difference if I use BgzipFastaAdapter as you recommended in your presentation at PAG2022. This is my first attempt at JBrowse. kind regards Hossein

…

On Tue, Nov 21, 2023 at 10:17 AM Colin Diesh ***@***.***> wrote: Hi there, How did you create the alignment? And how large was the resulting output file? Is it PAF format? JBrowse 2 does have to load the entire file into memory, and large alignments can exceed browser limits (1fffffe8->~532 megabytes) Even if they can be loaded, it may be slow. I would recommend possibly filtering out small alignments. You can also consider using a MCScan type alignment (just the protein sequences) instead of a whole genome alignment, as the datasets for MCScan are much smaller and much less tripped up by repetitive sequences ( https://github.com/tanghaibao/jcvi/wiki/MCscan-(Python-version)) Down the line, we are working on some things to make loading large whole genome alignments better by using indexing. If you are interested in trying that out as a beta branch I can try to help describe more. That work is here #3859 <#3859> — Reply to this email directly, view it on GitHub <#4069 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AWJHBCWCETOTPI3TSC7YRHLYFTHZBAVCNFSM6AAAAAA7UY4XZCVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TMMZSGYZTI> . You are receiving this because you authored the thread.Message ID: ***@***.*** com>

[cmdcolin]

are you using JBrowse desktop? can you provide a screenshot of the error?

[hugheb]

Hello Colin From your response, it seems JBrowse is for displaying alignment. If this is the case then I have got it all wrong. I was trying to do the alignment using an assembled reference and a draft genome. attached is the screenshot of the error message. thank you

…

On Tue, Nov 21, 2023 at 10:17 AM Colin Diesh ***@***.***> wrote: Hi there, How did you create the alignment? And how large was the resulting output file? Is it PAF format? JBrowse 2 does have to load the entire file into memory, and large alignments can exceed browser limits (1fffffe8->~532 megabytes) Even if they can be loaded, it may be slow. I would recommend possibly filtering out small alignments. You can also consider using a MCScan type alignment (just the protein sequences) instead of a whole genome alignment, as the datasets for MCScan are much smaller and much less tripped up by repetitive sequences ( https://github.com/tanghaibao/jcvi/wiki/MCscan-(Python-version)) Down the line, we are working on some things to make loading large whole genome alignments better by using indexing. If you are interested in trying that out as a beta branch I can try to help describe more. That work is here #3859 <#3859> — Reply to this email directly, view it on GitHub <#4069 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AWJHBCWCETOTPI3TSC7YRHLYFTHZBAVCNFSM6AAAAAA7UY4XZCVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TMMZSGYZTI> . You are receiving this because you authored the thread.Message ID: ***@***.*** com>

[cmdcolin]

I don't think the screenshot properly attached to your post

[hugheb]

attached is a PDF. hopefully it works this time thanks

…

On Wed, Nov 22, 2023 at 6:26 AM Colin Diesh ***@***.***> wrote: I don't think the screenshot properly attached to your post — Reply to this email directly, view it on GitHub <#4069 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AWJHBCW57V4YC63KWGCFYL3YFXVPDAVCNFSM6AAAAAA7UY4XZCVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TMNBRHAZTK> . You are receiving this because you authored the thread.Message ID: ***@***.*** com>

[cmdcolin]

sorry I think it doesn't work if you attach the file in an email reply. You can reply in the github page or send me a attachment via email at [email protected]