Wednesday, January 17, 2018

Gencove Review

Some of the apps Gencove offer
UPDATE: Gencove no longer accept uploads from other companies.

Gencove sells DNA tests for $59.99, but they also offer a free upload of your raw DNA data if you tested elsewhere. With the free upload, you get all the same options you do if you tested with them, namely an ethnicity report and matching with DNA relatives.

They also offer "apps", some from third parties. There is even a Promethease app for $10, though it would be cheaper and probably easier to just upload directly to Promethease (see a full review of Promethease here). There's also an app for GenePlaza, which was also previously reviewed here. Though the app is free, GenePlaza's reports each cost a small fee. Clicking on the GenePlaza app in Gencove merely takes you to GenePlaza's website. The other apps are free too, but they aren't particularly useful. They include:
  • Discover your microbiome - Bacteria and viruses that live in your mouth
  • My Genome - Info about your genomic data
  • Sleep - Are you a morning or evening person?
  • iobio.io - Compare your genome to ClinVar
  • YouGenomics India - Help improve genomics for South Asia
  • Gencove Mobile App - Compare results with friends on iOS or Android
  • Open Humans - Contribute to research and citizen science

When I tried Microbiome it simply said "Microbiome not available" with no explanation as to why, so that was totally useless.

My Genome is just that - it's where you find your raw DNA data. You can download your raw data, you can view which apps on Gencove you've given permission to access your data, and you can view and manage your consent to participate in research.

The Sleep app is interesting but the results claimed I'm a morning person, which I have never been. The app asks you a few questions about your sleep habits before showing your results and it did note "It seems that the genetic score and questionnaire results don’t match - an interesting outlier! That's probably because the genetics of sleep is not very well understood yet."

iobio
The iobio app loads your DNA to gene.iobio.io which is a little bit of a technical app that will tell you if you have any variants of certain medically related genes. For example, it includes a report on BRCA1 and BRCA2, which are genes associated with breast and ovarian cancer. Despite the technical looking nature of the site, it will tell you, in plain English, if you carry any variants of the genes included in the report or not. Hovering over each gene will pop up a brief summary of what it is associated it. Most of them are likely somewhat rare, since I had no variants for any of them. There are 40 in total: PTEN, BRCA1, BRCA2, TP53, STK11, MLH1, MSH2, MSH6, PMS2, APC, MUTYH, VHL, MEN1, RET, RB1, SDHD, SDHAF2, SDHC, SDHB, TSC1, TSC2, WT1, NF2, COL3A1, FBN1, TGFBR1, TGFBR2, SMAD3, ACTA2, MYH11, MYBPC3, MYH7, TNNT2, TNNI3, TPM1, MYL3, ACTC1, PRKAG2, GLA, MYL2, LMNA, RYR2, PKP2, DSP, DSC2, TMEM43, DSG2, KCNQ1, KCNH2, SCN5A, LDLR, APOB, PCSK9, RYR1, CACNA1S, ATP7B, BMPR1A, SMAD4, OTC. If you have reason to check on any of these and want a quick, free way to do this, this is a good option, but it can also be easily accessed independently of Gencove, just go to http://iobio.io, however, it's not very user friendly and I couldn't find a way to upload my data, so going through Gencove may actually be the better option.

YouGenomics India, recently renamed "Genavli Biotech", is a research project for South Asia, attempting to improve ethnicity reports for people with South Asian ancestry. Naturally, it wouldn't be useful for anyone who is not South Asian but if you are, you should look into this. As far as I can tell, Gencove's app simply links to the YouGenomics India website.

The Gencove Mobile App doesn't really offer anything that the website doesn't apart from some surveys which I presume are for research purposes. It allows your to see your ethnicity report and the unavailable microbiome report, and connect with or invite your friends. That's about it. 

The Open Humans app merely takes you to openhumans.org, which is an open research project. Gencove does not load your data there, so there's really no need to go through Gencove if you wish to participate in this project.

Gencove's populations for their ethnicity report
Most people will likely be most interested in the ethnicity report. There are 26 populations available, some of them are broad, large regions, while others only cover a small region. They include: Northern and Central Europe, Northern Italy, Northern British Isles, Southwestern Europe, Middle East, Eastern Mediterranean, Bengal, Central Africa, Eastern Africa, Northern Africa, Central Indian subcontinent, Southern Indian subcontinent, Oceania, Southeast Asia, Northeast Asia, Anatolia-Caucasus-Iranian Plateau, Central Asia, East Asia, North-central Asia, Northeast Europe, Scandinavia, Finland, Southern Africa, Western Africa, Ashkenazi Jewish, Americas. A map showing what these populations cover is shown left/above.

My personal results were not particular accurate, although I did note that if I added together all my results in populations probably associated with my Italian ancestry versus those from my Northwest European ancestry, the numbers were consistent with what most other companies say. Here are my Gencove results:

My Gencove ethnicity report
48% Northern and Central Europe
21% Northern Italy
15% Northern British Isles
7% Southwestern Europe
6% Middle East
3% Eastern Mediterranean

My Italian ancestry is southern, not northern, but if you add up the results for Northern Italy, Southwestern Europe, Middle East, and Eastern Mediterranean, you get 37%, which is very similar to the 36% from AncestryDNA and 38% from FTDNA. While my results in more specific regions may be all over the place across different companies, the divide between northern Europe and southern seems very distinct with me so when an ethnicity report is consistent with that, I know there's at least some reliability to it. 

Lastly, Gencove offer the "Relative Radar" which finds people you share DNA with. Unfortunately, there must not be very many testers/uploaders in their database because it found none for me so all I can say about it is that it seems to use a visual display, plotting relatives who share more DNA with you closer to your profile icon.

Conclusion: Since it's free, there's really no harm in checking out Gencove (unless you have concerns about research participation). Because some of their "apps" simply link to other sites, it looks like they offer more than they really do. The ethnicity report, sleep app, and iobio data were the only really interesting or useful options, but even with those, don't expect too much. I definitely wouldn't pay $59.99 to test with them, although the low price point in comparison with other testing companies may be appealing, you would get more out of your money by testing with AncestryDNA or 23andMe and then uploading to Gencove for free.

Monday, January 15, 2018

GPS Origins Upload Review

There are several different DNA tests available from HomeDNA, but this is a review of their "GPS Origins Algorithm" test which allows you to upload your raw DNA data from another company for an ethnic origins analysis. It costs $39 and as far as I can tell, it's pretty much the same results you'd get if you'd bought their "GPS Origins Ancestry Test" for $199. It does not include DNA matching with other testers or any kind of  health report.

Migration routes seem a little misleading, and generic
I found the results to be not very accurate, and even a little bit misleading. The information talks a lot about prehistoric and ancient origins, even though an autosomal SNP test doesn't generally go back that far. Other info says the test results "begin over 1,000 years ago" which is consistent with other autosomal ethnicity tests, which usually state the results represent about 500-1000 years ago. There's also a lot of information about Y and mtDNA which makes it sound like these include Y and mtDNA, even though it doesn't. It even has a map showing "migration routes", one in red and one in blue, which makes it look maternal and paternal. Not only is that impossible since I am female and don't have Y-DNA, but there is also a note: "Although both Migration Patterns represent your Maternal and Paternal DNA routes, we do not differentiate which route is maternal and which is paternal." So why make them blue and red? And why are there only two "routes"? I have several European origins, not only according to my tree, but also according to this ethnicity report. While GPS Origins do offer a Y test and an mtDNA test, they are not included in the autosomal DNA test.

Info on my Sardinian migration route when clicking on
one of the map markers (above)
The migration routes are suspicious because they don't offer much explanation as to how they determined these routes. From what I can tell, it seems like the "routes" are just generic info tossed up for people who get results in one part of Europe or another. For example, I got 11.7% in Sardinia, and my "blue" route shows it starting in Sardinia and moving north through Italy. Basically, the routes just seem to be providing generic info on the common migration pattern of Sardinians. The other "red" route appears to be from either my Fennoscandia results or Western Siberia. Again, I'm not very clear on how they chose these two routes, or why there's only two of them, apart from the fact that they seem to be trying to make them look like paternal and maternal results when they're not.

GPS Origins claims the migration route include "precision targeting—sometimes down to the village or town" which is a gross exaggeration. The map markers may look precise, but when you read the info attached to it, you can see it's not that specific.

The results in Sardinia, Fennoscandia, and Western Siberia are a little off to begin with. I'm getting 21.5% in Fennscandia, 11.7% in Sardinia, and 11.6% in Western Siberia. I do have Scandinavian roots in Norway, but it should only about 12.5%. The Sardinia results are probably coming from my Italian ancestry, but peaking in Sardinia doesn't seem right. As for Western Siberia, I have no ancestry there at all.

My ethnicity percentages
My full results are:

#1 Fennoscandia 21.5%
Origin: Peaks in the Iceland and Norway and declines in Finland, England, and France

#2 Sardinia 11.7%
Origin: Peaks in Sardinia and declines in weaker in Italy, Greece, Albania, and The Balkans

#3 Western Siberia 11.6%
Origin: Peaks in Krasnoyarsk Krai and declines towards east Russia

#4 Orkney Islands 11.5%
Origin: Peaks in the Orkney islands and declines in England, France, Germany, Belarus, and Poland

#5 Southern France 9.9%
Origin: Peaks in south France and declines in north France, England, Orkney islands, and Scandinavia

Why are they talking about Y and mtDNA in my autosomal
DNA results?
#6 Basque Country 9.3%
Origin: Peaks in France and Spain Basque regions and declines in Spain, France, and Germany

#7 Tuva 8.1%
Origin: Peaks in south Siberia (Russians: Tuvinian) and declines in North Mongolia

#8 Southeastern India 5.4%
Origin: Endemic to south eastern india with residues in Pakistan

#9 Northern India 5%
Origin: Peaks in North India (Dharkars, Kanjars) and declines in Pakistan

#10 The Southern Levant 3%
Origin: This gene pool is localized to Israel with residues in Syria

#11 Arabia 1.6%
Origin: Peaks in Saudi Arabia and Yemen and declines in Israel, Jordan, Iraq, and Egypt

#12 Northwestern Africa 1.1%
Origin: Peaks in Algeria and declines in Morocco and Tunisia

#13 Western South America 0.3%
Origin: Peaks in Peru, Mexico, and North America and declines in Eastern Russia

You can compare this with my results from other companies here.

The Legal Genealogist says GPS Origins is one not to bother with, and based on this I would agree. I definitely would discourage anyone from buying a test with them, considering it costs $199 compared to the more popular DNA companies charging $99 or less. You're basically paying twice the amount to get inferior ethnicity reports and no DNA matching with other testers. As for the $39 upload, it may be worth it for curiosity sake, but don't expect much from it. I found it very disappointing.

Here's another thorough review from another blogger.

Wednesday, January 3, 2018

The Importance of Primary Sources

This death certificate is only a primary source for the death
info. His parents names are actually wrong, and his birth year
is in conflict with records from when he was alive, suggesting
the informant, his son, may have gotten it wrong.
I often see people asking about which source is better for a certain fact or event and this is a good time to address the differences between a primary source and a secondary source. A primary source is a document which is recorded at the time of the event it's detailing. A secondary source is one that is detailing an event that occurred in the past, and therefore may be more likely to be incorrect. A primary or secondary source can also be a person, regarding whether or not that person was alive/witness to the event in question. So to understand the reliability of a record, we have to understand what it's a primary source for, and what it's not. Here's a quick rundown:
  • Birth records are the only primary source for a birth. This may include a birth announcement in the newspaper, but the further back you go, the less likely this becomes. Equally, the further back you go, the less likely that civil vital records were kept. Delayed birth certificates aren't a primary source, but may be the only record of a birth in existence. Also keep in mind that some places would fine individuals for reporting a birth too late, which means they may have lied about the birth date to avoid being fined.
  • Baptism records are only a primary source for the baptism, not the birth. However, if the baptism occurred only a few days after the birth, that's pretty much as good as a primary source for the birth too (if it recorded the birth date - do not assume the baptism and birth date are the same if both aren't recorded). Especially if there's no birth record in existence, a baptism record is likely as good as it's going to get. However, if the baptism took place years after the birth, maybe even months, that is not a primary source for the birth because enough time has passed for the actual birth details to have been remembered incorrectly.
  • Marriage records are only a primary source for the marriage. Particularly if the parents of the bride or groom were deceased, you can't be sure their names are correct. Be careful not to mistake a marriage bann, engagement announcement, or marriage license for the actual marriage.
  • Death records are only a primary source for the death. If it includes an address where the deceased was living at the time of death, then it's also a primary source for that. But it's a secondary source for the birth date and location, both because the document is normally recorded many years after the birth (unless it's an infant death), and because the informant for the death record is often someone who wasn't even alive or present at the time of the deceased's birth. It's also not a primary source for the parent's names or birth locations; it's very common for those details to be incorrect. Death records are usually a good source for the burial location, even though they are recorded before the burial takes place, and therefore that info theoretically could change before it happens.
  • Obituaries are generally considered a type of death record and therefore can be considered a primary source for the death if they are published within a few days of the death, as is typical. Excepting potential printing errors, of course (i.e. the informant may have provided the correct death information, but the typist misprinted it).
  • Gravestones aren't really a primary source for anything! At the most, they may be a primary source for the location of the burial, but I have seen gravestones erected for people before their death, who then actually wind up buried elsewhere. However, when this happens, there's usually a lack of at least a death date on the gravestone. It's also not a primary source for the date of the burial, since gravestones don't normally have the burial date listed on them. You might think it's a primary source for the death date, but gravestones often aren't created for weeks or even months after the death, plenty of time for people to remember the exact date incorrectly. 
  • Cemetery/burial records are only a primary source for the burial information. Unlike gravestones, these usually include the interment date and wouldn't exist unless the deceased was actually buried there.
  • Census records are only a primary source for data that was current at the time the census was taken, such as: residence, occupation, citizenship, literacy, etc. All other data that occurred prior to the census - birth/age, marriage, immigration, etc is secondary. Additionally, even things like the occupational data may be subject to the knowledge of the informant and could be incorrect. Also don't forget that in the US, pre-1880 censuses did not record relationships to the head of the household. While you can often surmise relationships based on the order in which people are listed, ages, and names, you can't be sure about them without other supporting documents to confirm. 
  • Family bibles may or may not be a primary source for any or all of the data within, depending on when each item of information was recorded and who recorded it. Unfortunately, there's generally no way to know for sure when the data was recorded, or who by. You can sometimes get an idea based on the handwriting and/or different types of pens used at different times. For example, you might note the birth info was recorded at a different time from the death info. But this still doesn't assure they were recorded at the time of those events. They could have each been recorded years after the fact, whenever the author (and we may not even know who the author was) got around to it.
  • Wills and Probates can contain a lot of valuable and reliable information, like the names of someone's children, the details of their estate/property, etc. But even though they are related to the death fact, they typically don't contain a date or location of the death, let alone a cause of death. Don't mistake the will or probate dates for the death date, but you can usually get a time frame for the death date - sometime after the will date, and before it was probated.
  • Lineage books are a secondary source for everything in them, since they are written after all the events took place. However, many lineage books use primary sources for at least some of their information. That doesn't necessarily mean the entire book is reliable, but the particular data coming from primary sources should be. Not all authors note their sources, but many do.
A gravestone with no dates - this person was
actually buried in a different cemetery (I
believe his parents erected gravestones for
their children in the family plot, but some of
them wound up choosing other cemeteries. This
is not typical in my experience.
Naturally, we do not always manage to find a primary source for each bit of information and that doesn't mean we can't use secondary sources. Even primary sources can be wrong sometimes, they are just much less likely to be so. We just have to work with what we have, and what exists, and understand what is more or less likely to be accurate. Having data from a secondary source doesn't mean we can't put that data or that source in our tree, it just means we should keep looking for better or additional resources to help confirm or deny it. Family trees are forever a work in progress and no one should assume that once a piece of information is put into a tree, it means you're confident it's accurate. The sources you cite in your tree should speak for themselves as to their reliability.

Judging which secondary source is more reliable for what type of conflicting data can be difficult, and we have to weigh when, how, and by who the data was recorded/provided. You may think it makes the most sense to go with a birth year that you find on most of the records for an individual, but what if all those records are from later in his life, or even after his death? A record from his childhood, closer to when he was born, and when his parents, who were there for the birth, were still alive and one of them may have been the informant may actually prove to be the more reliable source. Of course you can never know for sure, so it's also best to put all recorded facts in your tree as alternate data, but you still have to pick a default/preferred one. Hopefully, this has given you some things to consider when choosing a default/preferred fact to go with, and given you a good understanding of primary and secondary documents.