@James Fellows Yates has joined the channel
@Nikolay Oskolkov has joined the channel
@Jasmin Frangenberg has joined the channel
@Gunnar Neumann has joined the channel
@Martin Nathan has joined the channel
@Katerina Guschanski has joined the channel
@Viridiana Villa has joined the channel
Hi @channel you all showed interest in the SPAAM blog, so this channel is to discuss this.
as you can see above โ๏ธ the github repository is live, as is the website itself.
I want to ask you all to brainstorm ideas of what you would be willing to contribute (hint hint) or would like to see?
an โAsk SPAAMโ advice column! I guess this is also stackoverflow
Iโm just thinking like, what if someone wrote in with a question and we crowdsourced here and wrote up a summary of responses
Great idea! Also we can then speak to people who might be able to right a blog on the topic asked about
I ignored the bit about contributing but I would be happy to write a โsynthesis of what SPAAM people say to X questionโ especially if I can do it in Not Formal Academic Speak
That's great to hear! I already have one suggestion of a topic, and a person to do it even
I have a couple of other ideas too - will send when I get home
@James Fellows Yates perhaps too ambitious but I would be very interested in writing about comparing KrakenUniq vs Malt in terms of sensitivity/specificity on a synthetic data set like @Felix Key suggested. For example, one could gargamel simulate reads or simply chop reference genomes of a few bacteria, assign labels to the reads, run MALT and KrakenUniq on the reads, and compute ROC-curves to see which tool performs better (depending on the damage level?)
*Thread Reply:* I would definitely be interested in reading this!
@Nikolay Oskolkov I would actually rather other people curate/maintain the blog, so I'm going to volunteer @Shreya and @Ele to be the organisers of this. So they can ultimately decide ๐
But my ideas (with suggested authors)
Ah wait brb Maia is calling
I will accept any and all blog posts from Maia
Ok my ideas)
What exactly is the theta parameter in aDNA damage calculations? (suggested author @Nikolay Oskolkov) What are proof-reading enzymes, and why are they important for aDNA (suggested author @Christina Warinner) How to check for cross-contamination in ancient microbial genomics mappings using MultiVCFAnalyzer (suggested authors: @aidanva or @ร shild (Ash)) Who and how should I contact, when I want to consult with indigenous peoples about ancient metagenomics studies (suggested author @Justin Lund)
But you can ignore them all if you want
What momentum do you guys think would be good to keep the blog 'alive'? One a month?
*Thread Reply:* Once a month sounds ideal to me!
But I think your post idea would be fine @Nikolay Oskolkov ๐ (in the sense I would be interested to read it!)
Finally: let me know when you're ready for your first post and we can talk about how to add it ot the blog itself
(in the meantime I've added @Shreya and @Ele to the github team that administers the blog repository itself)
Sounds very good @James Fellows Yates, I can start preparing material for the theta-parameter post!
Had a couple of ideas from my own struggles 1) What is edit distance and why is it important; 2) internal barcodes (capitalist legacy); 3) Choosing a reference genome
*Thread Reply:* Maybe you could gather all the blog post ideas and put them as issues on the GitHub repository ๐
*Thread Reply:* Then you can assign people to it etc.
*Thread Reply:* I should've made you and @Shreya admins on it already
*Thread Reply:* Thanks James, will do this when I get a chance, feel like I have hype myself up for github. Maybe we should try and organise a meeting soon Shreya ๐
*Thread Reply:* Breadth vs depth of coverage (this is just turning into a list of things I google)
*Thread Reply:* you can do it!
*Thread Reply:* @Ele i've made a few issues of my own ideas as to what I meant, I think you can hype yourself up for that for the momemnt
https://github.com/SPAAM-community/spaam-blog/issues
(I'm happy to join the meeting to show you how adding the posts themselves would work :))
*Thread Reply:* Yes, letโs touch base and set up a plan @Ele and it would be awesome if @James Fellows Yates could pop in!
*Thread Reply:* I am out of town this week but back next Monday!
I suggest putting all the blog post ideas as github issues: https://github.com/SPAAM-community/spaam-blog/issues
(otherwise, as we are on a free Slack tier, we might lose the ideas if people talk to much
Inspired @ร shild (Ash): a clusterfuck of terms - usage of the word 'cluster' in ancient DNa
Btw @Shreya @Ele, @ร shild (Ash) inspired me so much I wrote something for the blog while sitting on a train with no internet yesterday. She's expanding it at the moment, so you'll have your first submission quite soon (it was an easy one to do, which is why I went for it ๐ฌ )
*Thread Reply:* Hope that's ok @Nikolay Oskolkov, don't want to jump ahead of you if you have something ready. I just thought maybe your ideas were more complicated so might take longer
*Thread Reply:* Ha-ha, no problems at all, please go ahead
Amazing!!!! Sorry been MIA, I have a progress review coming up and feeling STRESSED. Itโs on Monday (20), do you have time to meet later in that week? Iโm fairly flexible so open to time suggestions for Tue - Fri (21-24)
*Thread Reply:* Maybe you could make a doodle or something?
*Thread Reply:* Would be easier, especially with time zones for @Shreya
*Thread Reply:* Good luck with the review btw!
*Thread Reply:* Iโll make a Doodle! Good luck Ele!
*Thread Reply:* https://www.when2meet.com/?12877619-B8I9g
*Thread Reply:* Here you are! You can set your time zone up on the top left ๐ All things equal I would have a slight preference for Wednesday as I work from home that day so zooms are a little easier-- but Iโve marked off all my availability!
*Thread Reply:* Hey both ๐ I selected the slots you were both free. So Wed - Fri afternoon ๐ Really looking forward to this project
*Thread Reply:* How about Wednesday 9/22 at 9AM Chicago/ 2PM GMT?
*Thread Reply:* Works for me!
*Thread Reply:* Topic: SPAAM Blog Time: Sep 22, 2021 02:00 PM London
Join Zoom Meeting https://york-ac-uk.zoom.us/j/95063756661?pwd=RmhQcGRWZ0hrVzkwYnRTelZwOGFOQT09
Meeting ID: 950 6375 6661 Passcode: 4Vw93jrn
*Thread Reply:* Sent you both google calendar invites (York email for you @Ele as I guess York still uses google stuff? And Shreya's gmail)
*Thread Reply:* Hmm am having same problem with zoom link!
*Thread Reply:* @Shreya do you have a Chicago zoom thingy?
*Thread Reply:* Or we can meet on jitsi
*Thread Reply:* I do have zoom! Hang on let me send a link
@Christina Balentine has joined the channel
Markdown tips: https://github.com/adam-p/markdown-here/wiki/Markdown-Cheatsheet
@Ele @Shreya Unfortunately Ash is still overloaded at the moment, so it'll take a bit of time to get that blog post. I would suggest you start soliciting posts from others and I can help guide the submission itsefl. If you find who is willing already, let me know and I'll send some initial instructions for the github/markdown
By the way, the SPAAM blog could also eventually lead to a publication in JOSE (the Journal of Open Source Education) - https://jose.theoj.org/
That sounds very interesting
@Shreya @Ele I just had a thought. Originally I said we could try uploading to Zenodo automatically from GitHub for each post, but I realised that wouldn't be nice for making it citeable, because each post would include all the authors of all previous blog posts.
Instead I think it would make sense for each post to manually upload each post and get a unique DOI just for that post. That way you can also customise the author so it would just be the author of that specific post associated with the DOI.
We could do a manual upload at some point together if you wanted (I've done it once before)
*Thread Reply:* Iโve actually never used Zenodo, so the nuance here is a bit lost on me at this stage, but I think it makes sense!
*Thread Reply:* Ok. The auto thing is each time you make a 'release' of a repository, it would autoamtically upload the whole repository to Zenodo where it gets converted to an unmodifiable 'blob' file which is permanent so to say.
But the auto-method includes the entire changelog of the repository, rather just the 'latest' changes, so to say.
*Thread Reply:* But regardless, we can go through it toether ๐
*Thread Reply:* Could also try OSF.io instead ๐ค
JOSE would be really nice but it's very computational focused, so we would have to wait until we have a critical mass of computational posts before attempint it I thikn ๐
@Ele @Shreya we may have a first draft of a blog post for you next week
New bump for the spaam-blog channel ๐ Find info here: https://spaam-community.github.io/blog/2022/07/15/blog-welcome/
Potential crowdsourced blog idea
*Thread Reply:* If some one points me to how exactly the blog summary can be contributed, I can do this over the weekend!
*Thread Reply:* :partyparrot::poopparty::partydino: THANK YOU Kevin!! Whateverโs easiest for you! Probably the easiest would be writing it in a google doc and sharing it with me and @Ele at shreya23@uchicago.edu and eg715@york.ac.uk โฆ BUT if there is a different format you prefer, weโll work with that!
*Thread Reply:* @Gunnar Neumann, @Pooja Swali, @ร shild (Ash), @Meriam Guellil - I've paraphrased your comments as part of a blog post draft and credited you all+your institutions.
If you'd like to make any edits or changes please go ahead - you all have more experience in the field than I! If you'd rather have your quotation removed just say, no issue.
https://docs.google.com/document/d/1Vl1RAAtiNZzDPx1n1SC3aOQKg2pFHmKqMolRY7RBAIA/edit?usp=sharing
*Thread Reply:* I am fine with the quote. Maybe something to mention in other major factors influencing the effectiveness is fragment length. If you have a library with very small fragments you are just going to need a lot more sequencing because barely any reads will be assigned using most taxonomic identifiers. Min read length for classification is around 30-35bp usually, although there are exceptions. But then you could argue how accurate these exceptions are when they are based on fragments under 30bp. Anyways always good to have a look at your libraries first.
*Thread Reply:* Very good point! I've added the following: Similarly, libraries with a shorter overall fragment length profile will benefit from additional sequencing. The inherent challenge of identifying the origin of shorter DNA sequences will limit confidence in taxonomic or genome assignment. Libraries with median fragment lengths closer to the 30-35bp range - the typical (but not universal) minimum length cut-off - may require additional allocated reads.
*Thread Reply:* Hi @Kevin Daly, sorry for the late reaction and thanks to @Shreya for the reminder! quote looks good to me, i didnt contribute much to the conversation anyways. also text looks very good. thanks for writing this up!!
*Thread Reply:* @Shreya - is there interest in moving ahead with this?
*Thread Reply:* yes! @Ele and I were just talking about how great your post is!! Stay tuned (Iโm about to leave for a conference) and thank you again for writing it up!!
trying to keep all mentions of me as blog editor in this channel so I donโt lose them! @James Fellows Yates, @Nikolay Oskolkov, edit: @Maria Lopopolo, this sounds like a great three-person blog post!
*Thread Reply:* @Maria Lopopolo, are you finding the results of this discussion to be helpful? If so do you feel like writing a blog post summarizing what youโve found? cc @Ele
@Maria Lopopolo rather than Nico in this case (as the person who originally asked the question)
*Thread Reply:* ah yes, going to be honest i 100% have not had time to read the thread and just saw my tag
*Thread Reply:* And who is still working out what all of that means lol ๐
@Meriam Guellil has joined the channel
Hello! We have just released our next blog post ๐ This one is all about the number of sequencing reads needed to reliably screen for pathogens before enrichment. You can read all about it here:
https://spaam-community.github.io/blog/2023/04/07/blog-numreads/
This piece was written by @Kevin Daly and its the perfect example for how the SPAAM community can be used to crowd source answers to metagenomics questions ๐คฉ If you fancy doing the same then ask your question in <#C02DCKJ54JX|no-stupid-questions> and get in touch with @Shreya and @Ele
i hope it's of use to people, particularly those just entering the SPAAM community! thanks to everyone who actually contributed their knowledge (and @Ele & @Shreya for organizing it!)