The Biosphere

« first day (2827 days earlier) ← previous day next day → last day (1967 days later) »

James

09:32

@terdon Following on from [my answer](https://biology.stackexchange.com/a/86643/3553) about human variants:

I think I need to get the latest greatest gnomAD variant files. I am using a tool called VarMap to convert the gene coordinates to protein coordinates because it corrects the canonical mismatch between databases (up to 20% of proteins do not share the canonical transcript and proteins between Ensembl and UniProt).

I'll give you an update once I've wrangled the files!

terdon

@James Nice! Although I guess it doesn't make much sense to only look at canonical transcripts. The putative hyper invariant protein could well come from a secondary transcript.

I work for the folks behind VarSome, I'll talk to some colleagues to see if we have any resources that could help. I know we have all protein coding locations in a DB, so I might be able to extract something useful.

James

10:20

Of course, if there is a variant, we need to know its position in all/any isoforms it may appear in. The problem is an issue with, as always, mapping...

Let's say a gene is 30 nucleotides with an isoform of 15 nucleotides starting halfway through. The 30 nucleotide one is better annotated in terms of expression levels so that is canonical. However, the 5AA isoform has 10 structures!!! So, naturally, the protein folks pick that as the canonical.

We have a variant at the 2nd nucleotide. Without explicitly checking, we risk attributing our variant to the canonical protein isoform, even thoug…

And then comes the issue of mapping it to structure...

I'm fairly sure every group has solved this in their own way. But the VarMap method is probably the most comprehensive way of mapping from genes to structures with a lot of checks along the way. (I also work next to the folks that maintain it, so that makes bug-squashing easier!)

James

10:42

@terdon That would be super helpful! Thanks for checking :)

terdon

11:21

@james we've found a few.

varsome.com/gene/aatf

varsome.com/gene/SVBP

varsome.com/gene/DLGAP3

Those three have no known variants.

Ah no, hang on. My bad, there are many variants, just none with clinical information.

4 hours later…

Gen Test

15:00

Moderator @terdon . It's my first time answering here and first time on chat, so I really appreciate the feedback and help. I was wondering, if I'm having problems to complete the answer in a satisfactory way, I understand that a partial answer would still be ideal.

Gen Test

15:13

What would be an ideal way to ask for others to take it to the next step? As it's been a long time since I was in this field, and it may be beyond my abilities to complete the answer in a satisfactory way to the forum's standards.

terdon

@GenTest Hi, I'm afraid I'm not a moderator on Biology, but I am on other sites and I'm a relatively high-rep user here so perhaps I can help. Yes, partial answers can be useful, but it will depend on how partial they are :) If you can answer the main parts of the question, then please post. If not, maybe a comment would be better?

Ah, is this about your answer here?

Gen Test

Yes.

terdon

For what it's worth, I've already upvoted that.

Gen Test

I feel it's incomplete

but not sure how to express it

terdon

Hey @Chris, local mod help needed. Do you know who added the "more references" banner to Gen Test's answer above?

@GenTest It's kind of hard since the OP is obviously not entirely clear on the relationship between gene sequence and protein sequence. I think your answer does a decent job, but of course, it's still a guess.

So I assume whichever mod added the banner was hoping you could find an actual reference to back up your claim.

I would be very surprised if there is a protein with absolutely no known variation in the human population, but it is not impossible. For example, a short protein where the entire sequence is functional and, therefore, where any variant would be deleterious might exist.

Gen Test

15:18

ok, thanks. Just that I would like to add more calculations and forumlas from the neutral theory, but it's beyond my ability at this point. Was just hoping to find a way to ask if any other members might be able to complete it

terdon

But even then, your very valid point about the redundancy of the genetic code means you could still have synonymous variants.

Gen Test

or note that

terdon

@GenTest Or even just include a note: "Please feel free to edit and expand this if you can add more detail".

Gen Test

As I do agree with the flag

I shall do that! Thank you.

terdon

Or even, if you want to go the extra mile, make your post Community Wiki. That means you will no longer get reputation from it, but is also an invitation to the community to edit even more freely.

biology.stackexchange.com/help/privileges/community-wiki

892

Q: What are "Community Wiki" posts?

Some questions and answers are marked Community Wiki and are owned by a community wiki user: (in fact, this very question is one of them) Why have Community Wiki posts? How do Community Wiki posts work? How does a post become a Community Wiki post? How can the Community Wiki status be remove...

Gen Test

15:23

Thanks! I may just do that. I wasn't 100% sure if that function is something I can use.

Chris

16:18

@terdon It is from @theforestecologist

terdon

@Chris Might be a good idea to clarify what claims are unsupported. I was kind of surprised there was no explanatory comment by a mod.

Bryan Krause

16:38

@terdon It looks like it was in response to a flag, for which an explanatory comment already existed, and for which the OP has since improved the post.

terdon

@BryanKrause Ah, makes sense. Thanks for checking!

Bryan Krause

(removed the banner btw, I think in it's current form the contained claims are sufficiently referenced; though of course @GenTest further edits to add additional references are always acceptable :) )

terdon

It's an interesting question actually. I'll try and have a look at actual data over the weekend.

Gen Test

17:02

I think that an actual mathematical probability might be able to be derived using the "Infinite alleles model". But it's beyond my ability.

Always Confused

Q: Does this Rat psychological developmental mechanism also work in human?

In this epigenetics tutorial by University of Utah (Browser requires to support and allow adobe flash player); it is shown that in rat model those rat kids get little or no maternal care, whenn adult, develops aggression, lack of calmness etc. Whereas the rat kids got more maternal care and nurtu...