Has anyone had issues getting data back from ubiome? I have submitted close to 20 samples and would say that about 15% of them come back as being unable to process. I am not told what exactly the issue is. Any ideas on how to get better yield?
Hmm, my experience is that gut samples are very rarely unable to process. Skin, on the other hand, can be hit or miss. If you contact me directly I’ll talk to the uBiome lab and find out exactly what happened in your case.
My last uBiome was a 5 site and has come back unable to process.
How many samples in total have you had processed from ubiome?
Just 2.
My apologies for the delay. I really need to check the QS forums more frequently. I checked with the uBiome support team and learned that your kit has now been processed and you should be able to see the results.
It is sometimes the case that a sample is rejected on the first try. The most common reason is that the gene sequencer is unable to read enough DNA bases for it to meet the uBiome standards, in which case the sample is flagged for a later re-run, often after the lab people have a chance to see if it was a preventable failure that can be fixed next time.
Let me know if you have other questions about that sample, or feel free to show us your results if you’d like help analyzing them.
Thanks Richard. uBiome sent me a kit in place of this sample. How will I know in the future if a sample is going to be rerun? Should I assume that this sample does not pass their QC checks?
The results are here: https://github.com/isaacgerg/ubiome_longitudinal_analysis
I am curious to know about the large amount of haemophilus parainfluenzae. Any thoughts? Diversity looks pretty good and ubiome says this is one of my healthiest samples.
Thanks again for looking in to this.
Isaac
regarding Ubiome - how is there service doing now in terms of delays and failed samples ?
I am really interested to use their gut sequencing explorer service - but I can see from a quick google that there are a lot of people out there suffering v long or indefinite delays
some recent feedback would be great.
I find it takes roughly 8 weeks to get results, even longer with smart gut.
Unless things are messed up at the phyla level, I have not found anyone’s results usefule (I have 20+ and have looked at countless others).
I am part of the smart gut and it false alarms quite a bit. I’m putting together a technical paper now in response to ubiomes plos paper.
Thanks very much for the feedback.
I had seen a lot of negative stories from users saying they either didn’t get a kit, they got the kit but after submitting a sample they got a report saying insufficient material or some such error, or just didn’t get any results even 3 months later.
8 weeks is no problem if you know in advance.
out of interest - how do you mean smart gut false alarms quite a bit?
is it offering a diagnostic opinion on the results / levels of different bacteria?
I have only looked at explorer sample files so far and was considering a 3 x gut explorer package.
The smart gut report shows if a genus is high, low, or normal. These limits are defined as 0.5% and 99.5%-tiles respectively (the middle 99% are normal). They were determined from distributions of genus measured from a healthy cohort. These limits are defined in their plos paper.
The problem with their approach, and this leads to the false alarms, is that many of the distributions are very close to exponential with very small lambdas – many of them have none bacteria tested for. Another way to say it is that for a given genus, many folks have none of the bacteria. Therefore, if you would look at the smart gut report for the healthy cohort, they would report a low value of this bacteria.
A concrete example, look at Butyrivbrio crossotus. 88% of the healthy cohort has none of this bacteria. I used this in excel =COUNTIF(Table1[Butyrivibrio crossotus],”<0.000001")/COUNT(Table1[Butyrivibrio crossotus]). How do you define the percentiles of “normal” when 88% of the healthy cohort do not have this bacteria? How can you say the healthy cohort is “healthy” if you ran the smart gut report on them and 88% of them reported “low” for B. Crossotus?
From a machine learning point of view, they are doing kernel estimation and then creating a classifier. But, the kernels from which the data are generated contain many zero values so the classifier choice leads to a biased result giving the false alarms (this assumes the healthy cohort is in fact healthy – health was self-reported).
See paper here: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0176555
Data here: https://doi.org/10.1371/journal.pone.0176555.s003
All in all, I think ubiome is doing a good job of pushing the science. I wish more folks would get involved looking at this data critically because I think it benefits everyone.
Yep, I think I follow you, a basic problem of comparing various incomplete data sets/data sets with many zero values. They chose a way of doing that, but it has some issues. Thanks for the explanation.
Hi Isaac:
I took a look at the microbiome data that you provided a link to. (Many Thanks!)
-
I’m curious where you found this link as I was looking for it in the paper and could not find it.
-
When I looked at the data I noticed that a lot of the abundances are numbers greater than 1. Do you know what this means? Are they errors?
I am trying a few new approaches to analyzing this data. Any insights you could offer on this would be greatly appreciated.
Regards,
Jeff Lapides
jeffrey.lapides@gmail.com
That link is from Table S3 at the end of the paper. All numbers as percentages, i.e. range from 0 to 100.
What sort of new approaches are you trying?
Thanks Richard. Also responded via LinkedIn.
Jeff