• Welcome to Phoenix Rising!

    Created in 2008, Phoenix Rising is the largest and oldest forum dedicated to furthering the understanding of and finding treatments for complex chronic illnesses such as chronic fatigue syndrome (ME/CFS), fibromyalgia (FM), long COVID, postural orthostatic tachycardia syndrome (POTS), mast cell activation syndrome (MCAS), and allied diseases.

    To become a member, simply click the Register button at the top right.

Creating SNP panels - Feedback on appearance?

Messages
15,786
I'm working on SNP genotype panels relevant to ME/CFS patients (and might see if I can come up with a better methylation/detox one). My lovely fiance has the code mostly completed, but needs data as to how the result should look.

I want to keep things pretty simple. I don't like "+/-" = "positive/negative" because of the up/down association, which can get confused with up or down regulating genes. Same thing with the red and green colors.

So first of all I want to mark known up- or down- regulated genes separately. Blue for down-regulated (cooler, slower color) and red for up-regulated (warmer, faster color).

And since some panels are looking at risk-factors versus regulation, I want to use a non-color way to mark genes indicating a risk-factor. My current system is to put an exclamation mark after genotype results which are "positive" for that panel. This way it's easier to make it clear when a homozygous and heterozygous result are both "positive", instead of heterozygous results always looking semi-positive even when there's no risk from a heterozygous result.

I'm also debating whether to make non-positive results in normal print, while positive results are in bold, either for the allele or for the entire line (rs, allele, and gene). I've got samples of various possibilities in the following image files, so would like feedback on that, as well as any other problems spotted, or suggestions on making results easier to read and understand. The first two sections are all bold, the third section is bold for "positive" genotypes, and the 4th section is bold for the entire line where there's a positive genotype result. Please ignore any shaky or sloppy lines - those won't be in the final version.
output_sample.gif

Some text here to keep the pages separate.
output_sample_p2.gif

In case anyone's curious, these are the results that would be generated for the female sample file from 23andMe. The plan is to get the program producing these results up on the web somewhere soon.
 

LaurieL

Senior Member
Messages
447
Location
Midwest
Arrows instead of ! in a separate column denoting up or down-regulation. There are lots of brain-fog in those reading these results. As most are used to the +/+ denotation, as well as this is what they will encounter in other sources when trying to compare, you should really keep the +/+, etc. notations.

Gene first then the rs #.

Need a column for the risk allele after the rs #. I can't tell you how much time it has taken me personally to hunt these down.

+/- considered normal alleles should be marked with a *.

And again due to brain fog sufferers, only one column only.
 
Messages
15,786
Arrows instead of ! in a separate column denoting up or down-regulation. There are lots of brain-fog in those reading these results. As most are used to the +/+ denotation, as well as this is what they will encounter in other sources when trying to compare, you should really keep the +/+, etc. notations.
That's part of the reason I'm not doing +/- : the results really don't mean anything regarding up or down regulation as they're used by existing websites. But people assume it means the gene is down or up-regulated, even though it is only showing risk of (potentially) abnormal function or disease.
Gene first then the rs #.
I need to think about this one. On the one hand it's easier to see the gene you're looking for, but on the other hand it suggests that the results are for the gene. But they're really just for the rs#, which happens to be part of that gene - so getting an abnormal result for one rs# doesn't necessarily mean the gene is malfunctioning. If it's a good panel designed to look for malfunctioning genes then there should only be rs#'s relevant to the gene functioning, but that simply isn't the case with existing panels, and some of the ones I'm considering are like the example above - merely looking at rs#'s associated with risk, when the impact on the gene usually isn't known yet in the scientific literature.

Need a column for the risk allele after the rs #. I can't tell you how much time it has taken me personally to hunt these down.
The bold text and "!" is already showing which alleles that you have are the risk ones. It's another reason I'm not doing +/-. If the risk is only associated with having +/+, then there's no point in turning a +/- yellow or showing any other warning if there's no risk associated with it. Currently I have it set up so that multiple combinations can be tagged as risk. Usually it's one homozygous genotype, but sometimes it's the heterozygous which indicates greater risk - in which case the +/- system can be down-playing the problem :p
+/- considered normal alleles should be marked with a *.
Normal genotypes for the patient show up under genotypes, non-bolded and without the "!"
And again due to brain fog sufferers, only one column only.
I think you're right on this. I like having all the gene names grouped together, top to bottom, but the easier way to do it with a table in the software is left to right. So things are still grouped, but harder to see. I think a solution might be for each section to be in it's own table on the left or right side. If that still looks too crowded or confusing (or is too difficult from a programming perspective), then I'll go with a single column.

Oh, and here's the current output, as actually generated into a .pdf
 

Attachments

  • my.pdf
    7.8 KB · Views: 22

LaurieL

Senior Member
Messages
447
Location
Midwest
clicked on the my.pdf and it only displays a blank page? It may just be my old 95, limping on this computer.
 
Messages
15,786
Oh, and I just looked at the pdf again myself and realized how much zooming in is needed for the table to be readable. I think it'll definitely end up as one column, both to be more comprehensible and to have a big enough font to be readable when printed out!

Thanks LaurieL
 
Messages
15,786
clicked on the my.pdf and it only displays a blank page? It may just be my old 95, limping on this computer.
:eek:

I'll ask my expert (Mr Valentijn) if there's a compatibility issue between the modern .pdf files and archaic operating systems :D And if so, maybe we can include an option for generating .pdf files which will work on older operating systems.
 

roxie60

Senior Member
Messages
1,791
Location
Central Illinois, USA
Valentijn, maybe I'm dense today but I just took a look at one of the rs#'s for the panel on more severe FM. Your chart shows GG! as the high risk genotype for rs3794808 but in 23andMe they show C or T as the versions (I am CC) so I think this could be a point of confusion, where did you get a genotype of GG for this rs#? Is this the 'rule' of translating G=C and A=T? Also I thought all A and T's were risk alelles but I seem to see people indicating G and C versions can also be risk alelles, I wonder how the heck we are suppose to figure this all out if there seems to be no concensus on what it all means (feels like everyone speaking multiple foreign languages). I'm just curious to understand where you got the genotypes you are using in your chart.
 

roxie60

Senior Member
Messages
1,791
Location
Central Illinois, USA
Also rs165774 has gene ARVCF but in 23andME it is listed as gene COMT. I'm not an expert in any of this but just pointing out possible confusion if people are using your chart with their 23andMe results.
 

LaurieL

Senior Member
Messages
447
Location
Midwest
Archaic? Its downright historical, it takes all I have to have patience with it. :mad: This motherboard definitely has compatability issues.....I have one with it. :D
 

roxie60

Senior Member
Messages
1,791
Location
Central Illinois, USA
Still trying to be sure I am intrepreting your chart correctly: your chart lists a number of genotypes indicating they are predictive of FM/ME/CFS etc, but only a few are identified with ! in each panel. So what are we to make of the genotypes that do not have the !? I believe you are using the ! to help people know that some hetero are actually positive for the condition (not just possible indicator) but there are homozyg that also have the !. So I'm still trying to understand are all the genes in the panels indicators of the various conditions but if you have the genotypes with the ! you are even more likely to have the condition?? Sorry if I'm mucking up the works but I think you have a very promising tool here just need to be sure that at a glance everyone understands whet is being communicated. Also very much like the supporting references being included.
 

roxie60

Senior Member
Messages
1,791
Location
Central Illinois, USA
Normal genotypes for the patient show up under genotypes, non-bolded and without the "!"

I think this answers one of my questions regarding the other genotypes in the panel. If the non ! genotypes are normal (no indication of CFS/ME/FM) then why are they in a list for genotypes that have the potential to indicate the conditions, why not just incude the ! genotypes only (smaller list for now).?
 

NilaJones

Senior Member
Messages
647
I am so glad you are doing this :). Here are my thoughts on the samples.

I would leave out the explanation marks. Maybe it's my maths background, but to me they look like they are indicating something about the genome (like extra groups on the bases or summat).

Why not use the colors to highlight risk factors, red and blue up or down regulation? The heteros that are not considered risk factors, or where we are uncertain, could be a lighter shade, or bold but not color. Definitely bold the whole line.
 
Messages
15,786
Valentijn, maybe I'm dense today but I just took a look at one of the rs#'s for the panel on more severe FM. Your chart shows GG! as the high risk genotype for rs3794808 but in 23andMe they show C or T as the versions (I am CC) so I think this could be a point of confusion, where did you get a genotype of GG for this rs#?
Most of the available research uses A/G for rs3794808. So the program is translating the 23andMe T/C result for rs3794808 into the more useful A/G format.

Is this the 'rule' of translating G=C and A=T? Also I thought all A and T's were risk alelles but I seem to see people indicating G and C versions can also be risk alelles, I wonder how the heck we are suppose to figure this all out if there seems to be no concensus on what it all means (feels like everyone speaking multiple foreign languages). I'm just curious to understand where you got the genotypes you are using in your chart.
Yes, regarding the conversion. But no, and A and T aren't any riskier than other alleles in general. For a specific SNP, any one of the three combinations (such as CC, AC/CA, or AA) may be the riskiest one. Or none of the three might be associated with any risk, or two might be both associated with risk either to the same extent or one more than the other. You need to read the research to find out, and most alleles don't have any risk information discovered yet.
 
Messages
15,786
I think this answers one of my questions regarding the other genotypes in the panel. If the non ! genotypes are normal (no indication of CFS/ME/FM) then why are they in a list for genotypes that have the potential to indicate the conditions, why not just incude the ! genotypes only (smaller list for now).?
Because it might help give an idea of the scale of risk. By putting up 12 normal results and 3 risky results, it might indicate that there's less risk. But if only the three risky results are shown, someone might think they've got three out of three abnormal results.
 
Messages
15,786
Also rs165774 has gene ARVCF but in 23andME it is listed as gene COMT. I'm not an expert in any of this but just pointing out possible confusion if people are using your chart with their 23andMe results.
Thanks for catching that ... I checked the original research and I did indeed put the wrong gene :oops: It'll show COMT in the future :thumbsup:
 
Messages
15,786
I would leave out the explanation marks. Maybe it's my maths background, but to me they look like they are indicating something about the genome (like extra groups on the bases or summat).

Why not use the colors to highlight risk factors, red and blue up or down regulation? The heteros that are not considered risk factors, or where we are uncertain, could be a lighter shade, or bold but not color. Definitely bold the whole line.
I already have blue for down-regulation, and up-regulation will be red. But there is very little research showing whether a gene is up- or down-regulated as the result of specific SNP genotypes. And the up-, down-, or normal-regulation status is not always connected to the risk of a genotype, hence using colors and symbols separately - a symbol for risk, and red and blue (or white) for regulation.

Different shades are a good idea, but tricky to implement. Maybe hetero- is less risk than homo- for one SNP, but is it less risk than homo- for another SNP? So it creates a question as to which genotypes are bad enough to get the more alarming color. Instead of messing with multiple levels, I'm sticking with an "on or off" approach for now.

Any ideas for a different symbol to indicate a "positive" result for a test? I don't want +/- because of the confusion those create. Maybe an asterisk (*)? But then I expect to see a footnote or something :p Maybe a frowny face :( ? And I agree completely regarding bold for the whole line ... looks way better.
 
Messages
15,786
Archaic? Its downright historical, it takes all I have to have patience with it. :mad: This motherboard definitely has compatability issues.....I have one with it. :D
Mr Valentijn said putting it in other formats will be very easy. The .pdf will likely just be one option for people to view or download other formats.
 
Messages
15,786
I believe you are using the ! to help people know that some hetero are actually positive for the condition (not just possible indicator) but there are homozyg that also have the !. So I'm still trying to understand are all the genes in the panels indicators of the various conditions but if you have the genotypes with the ! you are even more likely to have the condition??
The research I've been looking at usually indicates which genotypes are a better indicator of risk. They usually present results for all three genotypes (AA, AB/BA, BB, of each SNP they examine), so it's clear when both AA and AB/BA are at greater risk, or if only BB is at greater risk, or if only AB/BA is at greater risk.

That's one reason I'm moving away from the systems where heterozygous is always yellow and +/-. Usually it's known if that heterozygous genotype is a risk or not, and there's no reason to leave it looking so ambiguous.
 
Messages
15,786
Footnotes may not be a bad idea? Will save additional time spent researching for some folks.
Sources for the SNPs used are already cited, which should help a lot.

I'm interested in putting a bit more information possibly. With some SNPs it's known how people will react to certain drugs, etc, which is useful, and it would be nice to show the level of risk. But calculating the actual level of risk is often not done in the papers, unless I can somehow figure that out based on p values. Anyhow, it's something that can be added later possibly, once the basics are working.

Another possibility is an automated interpretation guide. Again, a future project that might or might not happen :p