When I made this video, I was a rising senior in high school. Now I'm in my second semester of college and I am researching at a genetics lab again (Lachance lab at Georgia Tech). Hopefully in the future I'll be able to produce more videos like this explaining population genetics concepts! Glad that it has helped everyone so far!
I second this comment recorded before : "Excelent,very informative. You have managed to explain a concept which university lecturers struggle to explain in 90 minutes. Great work." Good job.
Brilliant. I did/do graduate work in psychology and neuroscience and GWAS/TWAS is really important for heritable disorders and understanding epigenetics. This simplifies what many university faculty fails.
I third this comment recorded before: "I second this comment recorded before : "Excelent,very informative. You have managed to explain a concept which university lecturers struggle to explain in 90 minutes. Great work." Good job." Good work. Great job.
Before I watched this video I lived in a one-bedroom apartment and made minimum wage. Thanks to Nuno and his video I made 6 billion dollars last year and live in a mansion in Beverly Hills. Thanks Nuno!
Amazing video! The only comment I'd make is @6:00 that a Bonferroni correction is pretty conservative in this scenario. Bonferroni is great when the SNPs (or any data points, for that matter) are independent. Given that numerous SNPs are in LD with neighboring SNPs, they are not completely independent from one another. Many would argue that it's reasonable to use a somewhat less stringent correction. But better stringent than too lenient! :)
Thank you for this video, I will be watching it probably 2 more times for full comprehension (lol), but it is very well laid out and informative. I wish you great success in your studies and beyond :)
I am following this account right away. I love how you explained GWAS. I joined a genetics lab and the terminologies used confuses me a lot. This will surely give me a good head start.
I am actually working on cancer genetics research at Tulane. I wanted an easier way to explain GWAS to public health collaborators with limited knowledge on genetics. Your video is great to help me organize my thoughts!!!
Excellent excellleenntttt video!!!!!!! I have been struggling with understanding what GWAS is an its purpose but you did an AMAZING job with explaining in under 10min
excellent and surprised there is no other nags methodology, would be perfect to hear from such understanding and informative way from you this helpe me a lot so associate epigenetics and analysis
Thank you for this very well explained video ! Your presentation and your explanations will be very useful for me, thanks to you, I understood the principle very easily. Very good work ! I hope I could present my PhD as you explain your videos :)
Very informative video. Would love to see more on GWAS. I've been looking for your socials to contact you and ask you to post more on this, as I'm very eager to learn about it. ( My major in Bachelors is Plant Breeding and Genetics) Also, i did not find anything to contact u and ask u for posting more.
I'm glad you liked the video! I am still researching in the field of genetics (soon going to be doing a PhD in bioinformatics), so I may do more videos in the future if I can fit them in my schedule and believe it'll add value to the field.
Is the Bonferroni correction really always the method of choice for GWAS? I learnt that this correction is unnecessarily aggressive for many purposes and that you can use other forms of correction in most cases. So is GWAS a case where Bonferroni actually is the best method and you should not use another one?
From memory, I cannot recall seeing a method of multiple testing correction used that wasn't Bonferroni. More specifically, a threshold of p < 5E-8 is almost always used regardless of the actual number of SNPs tested. You're right that the Bonferroni correction is overly aggressive in the context of GWAS, because SNPs are often in linkage disequilibrium (LD) with each other, and as such, the effective number of independent association tests being run (which is what Bonferroni attempts to control for) is less than the number of SNPs. A less stringent correction method is the false discovery rate (FDR; also known as Benjamini-Hochberg procedure) which I have used in a lot of analyses, but not GWAS specifically. Ultimately, the choice of correction depends on what you are trying to minimize, false positives or false negatives. It seems that typically, population genetics researchers prefer minimizing the former.
You may find this paper helpful: www.ncbi.nlm.nih.gov/pmc/articles/PMC6001694/ That paper is how I learned to use PLINK (as well as by receiving help from a graduate student at my lab)
How is the genome typically collected and transferred as data onto software? Is it a blood or saliva collection that is analysed through microscopes then somehow transferred as data or collected by different methods?
That's something beyond what I know, since usually the data collection and data analysis are separate steps completed by different teams. When I worked on GWASs last summer, I used data from UKBioBank, and you can read a little bit about their data collection on their Wikipedia article: en.wikipedia.org/wiki/UK_Biobank I'd imagine the DNA data collection depends on the source you're getting the data from, but at least UKBioBank seems to have gotten their data from extracting DNA from "blood and urine samples"
Generally, dominant alleles are written in capital letters while recessive alleles are written in small letters. Can you specify which one is the dominant allele in your case(A or T)?
From my understanding, that information is usually present on a spreadsheet that gives information on the genes in your dataset. Unfortunately, I have not worked with PLINK since making this video, so I don't remember the exact file type name. But in that spreadsheet, if you got your data from a good source, you will find that it will say what the dominant allele is. Either way, for the GWAS explained in this video, it does not matter much which one is the dominant and which one is the recessive, since the GWAS just discovers which SNPs are associated with the phenotype. And keep in mind, in this video, A and T are nucleotides, not alleles necessary. If you want to determine causality (eg: a T allele increases the phenotype measurement, but an A allele decreases it), I believe PLINK can tell you that information by looking at the effect size it found for that SNP (if it's positive or negative, and the magnitude tells you how big of an effect that SNP has). However, I do not have much experience with that. Hope this answered your question.
I would not call them associations between variations in our "genetic code". The rough definition of genetic code is the set of rules used by living cells to translate information encoded within genetic material into proteins, and in a nutshell refers to the correspondence between trinucleotides and the way they match with t-RNA-linked aminoacids. I would rather says association between variations in our "genes, gene sequences, etc., but not in the code as such. The code is universal across species, with certain exceptions.
Hi! I currently have a genetics presentation on GWAS, I was wondering what program you used to animate this video? As I need to produce something similar (though for a very specific paper so I'm not copying any of your ideas, merely looking for animation apps haha). Thanks!
I used Adobe Illustrator to make some of the graphics (like the background, the cell, the chromosomes, and a few others), Powerpoint for some of the icons (like the man/woman, the globe, etc), and I used Adobe After Effects to animate everything. Just as a warning, it takes a lot more time to animate a presentation with a video editing software as opposed to an actual presentation tool like Powerpoint or Google Slides. If you do not have much time to submit your presentation and don't care too much about seamless transitions and editing quality, I would not recommend using a video editing software. Let me know if you have more questions
@@NunoCarvalho Ah currently going through the same situation. I have a presentation in few hours about GWAS in maize. What maize? currently I am struggling with the basic concept of GWAS. Idk I would be able to give it or not. Its soooo complex in maize.
Very nice video Nuno! Let me just make a correction: in the beegining of the video, when you say "variation in the genetic code", this is conceptually wrong. The genetic code is just one, the one that stablishes which aminoacid is coded by which trinucleotide, and this is not variable. What you really meant was the variation in the genome ;)
Yes I have since realized this error, thank you! I was unaware that genetic code referred to a specific concept in genetics, I was indeed just using to refer to the genome.
Great video！I have a question: SNPs can be used as a marker for the study. It means the SNP or near loci associated with the disease. It requires a database of SNP should have SNPs linked to the phenotype. Could you explain the SNP array, like "Genome-Wide SNP Array ", How to make sure they have the SNP linked to the phenotype?
There are many different arrays available on the market that differ in SNP density. No matter the density, most SNP arrays are spread out to cover as much of the genome as possible. If you have a specific research question, you can often find specialized arrays that genotype SNPs that have previously been determined to be associated with your area of interest. And if you want greater coverage after the fact you can try to impute SNPs (leverages linkage disequilibrium to calculate the probability of a SNP (that wasn't on your array) being a specific allele given the neighboring SNP's allele (which you do know) and the ancestral background of the subject who was genotyped).
When I made this video, I was a rising senior in high school. Now I'm in my second semester of college and I am researching at a genetics lab again (Lachance lab at Georgia Tech). Hopefully in the future I'll be able to produce more videos like this explaining population genetics concepts! Glad that it has helped everyone so far!
Excelent,very informative. You have managed to explain a concept which university lecturers struggle to explain in 90 minutes. Great work.
Thank you!
I second this comment recorded before : "Excelent,very informative. You have managed to explain a concept which university lecturers struggle to explain in 90 minutes. Great work." Good job.
Brilliant. I did/do graduate work in psychology and neuroscience and GWAS/TWAS is really important for heritable disorders and understanding epigenetics. This simplifies what many university faculty fails.
Fantastic! This is the best description of GWAS yet! Thank you
I third this comment recorded before: "I second this comment recorded before : "Excelent,very informative. You have managed to explain a concept which university lecturers struggle to explain in 90 minutes. Great work." Good job." Good work. Great job.
Honestly man, this is the best explanation I've come across. Thanks so much
I'm new to the field of genetics & found your video to be SUPER helpful! Would love to see more!
Glad it helped you!
Before I watched this video I lived in a one-bedroom apartment and made minimum wage. Thanks to Nuno and his video I made 6 billion dollars last year and live in a mansion in Beverly Hills. Thanks Nuno!
Amazing video! The only comment I'd make is @6:00 that a Bonferroni correction is pretty conservative in this scenario. Bonferroni is great when the SNPs (or any data points, for that matter) are independent. Given that numerous SNPs are in LD with neighboring SNPs, they are not completely independent from one another. Many would argue that it's reasonable to use a somewhat less stringent correction. But better stringent than too lenient! :)
This video literally changed my life!! incredible work
Super clear explanation. I need more video like this for different topics.
Thank you for this video, I will be watching it probably 2 more times for full comprehension (lol), but it is very well laid out and informative. I wish you great success in your studies and beyond :)
This is such a great and informative video! Definitely beats reading a dozen publications on GWAS and SNPs.
I am following this account right away. I love how you explained GWAS. I joined a genetics lab and the terminologies used confuses me a lot. This will surely give me a good head start.
Glad to hear that it's helped you
I am actually working on cancer genetics research at Tulane. I wanted an easier way to explain GWAS to public health collaborators with limited knowledge on genetics. Your video is great to help me organize my thoughts!!!
Excellent excellleenntttt video!!!!!!! I have been struggling with understanding what GWAS is an its purpose but you did an AMAZING job with explaining in under 10min
This is very helpful. You can sum up hours of lecture into 7 minutes video. Great work! Thanks
excellent and surprised there is no other nags methodology, would be perfect to hear from such understanding and informative way from you this helpe me a lot so associate epigenetics and analysis
Thank you for this, I’m researching resistance and new to the field this definitely helped
This really helped me hardening my understanding while reading Robert Plomin's "Blueprint". Thank you!
This helped me with my masters thesis, thank you!
Heyo nice explanations and video! this is the guy from BMES btw :)
Thank you! Best of luck with college applications!
Best video on GWAS!! Thanks a million. Pls do on LOD
I am new to GWAS.. Really very nice and well explained video.. thanks
Masterfully thorough yet concise explanation, thank you so much!!!
Kudos to you! It is very helpful to generally comprehend the concept of GWAS.
This is a great video, Nuno.
Thank you so much for making this video! Clear explanation, and good explanation of linkage of the biology and statistical components of GWAS!
Great! please discuss Linkage Disequilibrium
Such a comprehensive and informative video.
great work! Your explanation is so intuitive and easy to understand. Thank you!!
Very clear and thorough explanation! Thx!
I wish I could give multiple likes in this video! AWESOME!
What a great video. Thank you for this explanation
Nice and crisp explanation.
Perfect Explanation!
Perfect! Your video is very useful, thank you
Thanks for your great video. I expect the next video too.
thank you very much for making this video! you make it easy to understand! really help me a lot!
Thank you, it made things so much more clear !
Wow i finally understand, thank you!!
Thank you for this very well explained video ! Your presentation and your explanations will be very useful for me, thanks to you, I understood the principle very easily. Very good work ! I hope I could present my PhD as you explain your videos :)
Very nice video...super informative
Very informative video. Would love to see more on GWAS. I've been looking for your socials to contact you and ask you to post more on this, as I'm very eager to learn about it. ( My major in Bachelors is Plant Breeding and Genetics) Also, i did not find anything to contact u and ask u for posting more.
I'm glad you liked the video! I am still researching in the field of genetics (soon going to be doing a PhD in bioinformatics), so I may do more videos in the future if I can fit them in my schedule and believe it'll add value to the field.
I liked this explanation, it makes clear some of the data analysis done in GWAS. Can you explain LOD score please?
Great work.
Thank you for making this video.
Very nicely explained
Well done! This video was really helpful!
Thank you for the explanation, it's really helpful!
Is the Bonferroni correction really always the method of choice for GWAS? I learnt that this correction is unnecessarily aggressive for many purposes and that you can use other forms of correction in most cases. So is GWAS a case where Bonferroni actually is the best method and you should not use another one?
From memory, I cannot recall seeing a method of multiple testing correction used that wasn't Bonferroni. More specifically, a threshold of p < 5E-8 is almost always used regardless of the actual number of SNPs tested.
You're right that the Bonferroni correction is overly aggressive in the context of GWAS, because SNPs are often in linkage disequilibrium (LD) with each other, and as such, the effective number of independent association tests being run (which is what Bonferroni attempts to control for) is less than the number of SNPs.
A less stringent correction method is the false discovery rate (FDR; also known as Benjamini-Hochberg procedure) which I have used in a lot of analyses, but not GWAS specifically.
Ultimately, the choice of correction depends on what you are trying to minimize, false positives or false negatives. It seems that typically, population genetics researchers prefer minimizing the former.
Great video!
please help if i want to find obesity SNPs for the specific population
wonderful video--great job!
What kind of input genomic material (RNA/DNA etc) do you use for genotyping?
Awesome.....I am impressed 👏👏 could you please explain how to use plink.
You may find this paper helpful: www.ncbi.nlm.nih.gov/pmc/articles/PMC6001694/
That paper is how I learned to use PLINK (as well as by receiving help from a graduate student at my lab)
How is the genome typically collected and transferred as data onto software? Is it a blood or saliva collection that is analysed through microscopes then somehow transferred as data or collected by different methods?
That's something beyond what I know, since usually the data collection and data analysis are separate steps completed by different teams. When I worked on GWASs last summer, I used data from UKBioBank, and you can read a little bit about their data collection on their Wikipedia article: en.wikipedia.org/wiki/UK_Biobank
I'd imagine the DNA data collection depends on the source you're getting the data from, but at least UKBioBank seems to have gotten their data from extracting DNA from "blood and urine samples"
Generally, dominant alleles are written in capital letters while recessive alleles are written in small letters. Can you specify which one is the dominant allele in your case(A or T)?
From my understanding, that information is usually present on a spreadsheet that gives information on the genes in your dataset. Unfortunately, I have not worked with PLINK since making this video, so I don't remember the exact file type name. But in that spreadsheet, if you got your data from a good source, you will find that it will say what the dominant allele is. Either way, for the GWAS explained in this video, it does not matter much which one is the dominant and which one is the recessive, since the GWAS just discovers which SNPs are associated with the phenotype. And keep in mind, in this video, A and T are nucleotides, not alleles necessary.
If you want to determine causality (eg: a T allele increases the phenotype measurement, but an A allele decreases it), I believe PLINK can tell you that information by looking at the effect size it found for that SNP (if it's positive or negative, and the magnitude tells you how big of an effect that SNP has). However, I do not have much experience with that.
Hope this answered your question.
I can't thank you enough, very useful
Can you do eQTL/QTL analysis next please???
Great video.
I would not call them associations between variations in our "genetic code". The rough definition of genetic code is the set of rules used by living cells to translate information encoded within genetic material into proteins, and in a nutshell refers to the correspondence between trinucleotides and the way they match with t-RNA-linked aminoacids. I would rather says association between variations in our "genes, gene sequences, etc., but not in the code as such. The code is universal across species, with certain exceptions.
God bless you
Hi! I currently have a genetics presentation on GWAS, I was wondering what program you used to animate this video? As I need to produce something similar (though for a very specific paper so I'm not copying any of your ideas, merely looking for animation apps haha). Thanks!
I used Adobe Illustrator to make some of the graphics (like the background, the cell, the chromosomes, and a few others), Powerpoint for some of the icons (like the man/woman, the globe, etc), and I used Adobe After Effects to animate everything.
Just as a warning, it takes a lot more time to animate a presentation with a video editing software as opposed to an actual presentation tool like Powerpoint or Google Slides. If you do not have much time to submit your presentation and don't care too much about seamless transitions and editing quality, I would not recommend using a video editing software. Let me know if you have more questions
@@NunoCarvalho Ah currently going through the same situation. I have a presentation in few hours about GWAS in maize. What maize? currently I am struggling with the basic concept of GWAS. Idk I would be able to give it or not. Its soooo complex in maize.
@@eatlikethelocals93 I hope your presentation went well!
Can u plz tell me what is meant by 'casual variants'?
Causal variants not casual - ie variants that cause the phenotype
Very nice video Nuno! Let me just make a correction: in the beegining of the video, when you say "variation in the genetic code", this is conceptually wrong. The genetic code is just one, the one that stablishes which aminoacid is coded by which trinucleotide, and this is not variable. What you really meant was the variation in the genome ;)
Yes I have since realized this error, thank you! I was unaware that genetic code referred to a specific concept in genetics, I was indeed just using to refer to the genome.
Good job
Nice video thanx
Great video！I have a question: SNPs can be used as a marker for the study. It means the SNP or near loci associated with the disease. It requires a database of SNP should have SNPs linked to the phenotype. Could you explain the SNP array, like "Genome-Wide SNP Array ", How to make sure they have the SNP linked to the phenotype?
There are many different arrays available on the market that differ in SNP density. No matter the density, most SNP arrays are spread out to cover as much of the genome as possible. If you have a specific research question, you can often find specialized arrays that genotype SNPs that have previously been determined to be associated with your area of interest. And if you want greater coverage after the fact you can try to impute SNPs (leverages linkage disequilibrium to calculate the probability of a SNP (that wasn't on your array) being a specific allele given the neighboring SNP's allele (which you do know) and the ancestral background of the subject who was genotyped).
Very clear, easy to follow, but rigorous.
If your family name is "Carvalho", your ancestors are probably from Portugal.
Keep on the dood work
Jorge Carvalho
Sim, sou de Portugal :) Agora vivo nos Estados Unidos.
@@NunoCarvalho Best Regards
W video
Sana helal olsun
Good lesson but a little fast.
Vídeo interessante, não precisavas de massacrar o teu nome em Inglês haha