Great, if you get into different design of experiments and points, it would be better. I have the same difficulties with my students in our lab, as they do not know what to repeat and do not know the reason for their repetition in a set of experiments
Thank you very much for all your videos. I'm currently preparing to enter in an international enterprise thanks to your videos. I've already bought your book but I am going to also buy something else ! Your content is so precious that I have to give you my money for your help. Just one question : In the case of a biological replicates, as we measure only one time for each people, measure wouldn't be accurate compared to technical replicates. Isn't it a problem ?
Usually the goal is to get an estimate of the overall population not an estimate of the individuals. The technical replicates will only help give us a good estimate of the individuals. The only way to get a better estimate of the overall population is to measure more individuals, rather than the same individual multiple times. To see why, see: th-cam.com/video/gKnfP2_Xdpo/w-d-xo.html
If I mixed both biological replicates and technical replicates in the experiments, how should the sample size be determined when comparing the differences between the experimental group and the control group? Should the number of biological replicates be used, or should it be multiplied by the number of technical replicates? Thank you.
Thanks for the awesome video! I did have a question about technical replicate though. If at day 1, I run an experiment on a sample where I measure the absorbance (measure the absorbance 3 times), is this considered a technical replicate? Or would an example like at day 1, measure the absorbance of a sample 3 times, then day 2 and 3, do the same with the exact sample from day 1, is this considered technical triplicate? So essentially, would measuring the absorbance 3 times be considered technical triplicate in this case or would total of 9 measurements be technical
If you are using the same materials on day 1, then those are technical replicates. If you then repeat the same thing, with the same materials, on day 2, then you have a mixture of technical replicates plus a measurement of something called a "batch effect". For example, if I did the experiment on day 1 and you did the experiment on day 2, even though we both used the same material, it might be that you are better at the experiment, and I am more careless, so my values will have more variation than yours and that means we would have a "batch effect" based on who did the experiment.
@@statquest makes sense thank u! Yes it was same material in my case since all I'm doing is making the reaction mixture and pipetting same volume of that reaction mixture into 3 wells. So it is a technical replicate!
@@chw512 Note: Even if it is you doing the experiment on both days, there can still be a batch effect (maybe you sleep better before one batch of experiments etc.)
how do you combine technical and biological replicate variance if both are of the same order.....or if tech replicate variance is large and could affect the comparisons in the biloical replicates the object if the study
You have to create a complicated design matrix to accommodate both types of replicates. To learn more, see my series of videos on linear regression: th-cam.com/play/PLblh5JKOoLUIzaEkCLIUxQFjPIlapw8nU.html
Hi Josh, Quick Question: For cell lines, if I were to use only one particular cell line and repeat it. If I have done this experiment 3 times, as compared to having 3 different measurements in 1 experiment. Would that be considered biological replicates for repeating the experiment 3 times?
May I help you with this? Look, you are looking for the logic that you repeat a measurement. If you are going to do a cell experiment single cell line, you use technical replicates, which is the repetition in one experiment or repeating the experiment three times. All you are doing is that you measure something in a cell line. in the first, you measure it in three collection of the cells from the same cell line, in the latter you measure it along with the variation in different set of experiments. If you do the same experiment in multiple cell line that will be your biological replicates. Technical replicate is related to the techinque profile such as precision, data variation in the technique, etc.
Hi StatQuest, thanks for the video. It confuses me a bit because in the examples you give of technical replication, seems like an example of pseudoreplication, due to the samples are not independent of each other. Am I right? if not, why? Thank you again.
In this case pseudoreplication is the same thing as a technical replicates since the measurements are correlated. However, the goal in this case (for technical replicates) is to measure the technical error, and we do that by removing the biological variation - thus, we do the same experiment from the same biological sample.
Hi Josh. Thanks for your video. I am still confused about my case. Recently, I am collecting sample for RNAseq. My sample is pig oocyte. Each week, I can get 10 ovaries from 5 individual pigs (one pig contains two ovaries). Normally, one ovary contains 10-20 healthy oocytes so I can get ~150 oocytes I want each week. For establishing the RNA library, 30 oocytes is required. If I want to have 4 biological repeats (4 repeats need 120 oocytes), can I finish it within one week? Or I should collect 30 oocytes each week and finish sampling 4 weeks later?
@@statquest I don't understand. If I have a bacteria, and I have RNASeq data of 3 of it's strains under same condition, are these biological replicates?
Support StatQuest by buying my book The StatQuest Illustrated Guide to Machine Learning or a Study Guide or Merch!!! statquest.org/statquest-store/
Every few years or so I stumble again on statquest while googling something, and remember your awesomeness! Thank you Josh for this amazing series
Thank you!
Extremely helpful! Thanks, dude! I was wondering about this in one of my experiment.
Thanks StatQuest. Love the way u explain concepts.
My pleasure!
StatQuest is Best! Very, very practical!
Thank you!
I love StatQuest.
Thanks, dude! It's extremely clear for Taiwanese student like me!
Your videos are super awesome!!! I have learned a lot from them. Thanks
Hooray! :)
This dude is thankful
Bam! :)
Hi Josh, thanks for the video. How does one deal with standard deviation for multiple biological replicates each with multiple technical replicates?
Very good explanation
Thanks! :)
There is a state quest for anything is there. Just great!
BAM! :)
Way easier than google's explanation.
Thanks!
Your way of explaining is immaculate for neophyte learners...Thanks a lot
Thank you! :)
@@statquest Welcome 👍
extremely helpful, thanks
Glad it helped!
Great, if you get into different design of experiments and points, it would be better. I have the same difficulties with my students in our lab, as they do not know what to repeat and do not know the reason for their repetition in a set of experiments
Thank you very much for all your videos. I'm currently preparing to enter in an international enterprise thanks to your videos. I've already bought your book but I am going to also buy something else ! Your content is so precious that I have to give you my money for your help.
Just one question : In the case of a biological replicates, as we measure only one time for each people, measure wouldn't be accurate compared to technical replicates. Isn't it a problem ?
Usually the goal is to get an estimate of the overall population not an estimate of the individuals. The technical replicates will only help give us a good estimate of the individuals. The only way to get a better estimate of the overall population is to measure more individuals, rather than the same individual multiple times. To see why, see: th-cam.com/video/gKnfP2_Xdpo/w-d-xo.html
If I mixed both biological replicates and technical replicates in the experiments, how should the sample size be determined when comparing the differences between the experimental group and the control group? Should the number of biological replicates be used, or should it be multiplied by the number of technical replicates? Thank you.
I believe you should just use the biological replicates.
Thank you! Very helpful :)
bam! :)
Thanks for the awesome video! I did have a question about technical replicate though. If at day 1, I run an experiment on a sample where I measure the absorbance (measure the absorbance 3 times), is this considered a technical replicate? Or would an example like at day 1, measure the absorbance of a sample 3 times, then day 2 and 3, do the same with the exact sample from day 1, is this considered technical triplicate? So essentially, would measuring the absorbance 3 times be considered technical triplicate in this case or would total of 9 measurements be technical
If you are using the same materials on day 1, then those are technical replicates. If you then repeat the same thing, with the same materials, on day 2, then you have a mixture of technical replicates plus a measurement of something called a "batch effect". For example, if I did the experiment on day 1 and you did the experiment on day 2, even though we both used the same material, it might be that you are better at the experiment, and I am more careless, so my values will have more variation than yours and that means we would have a "batch effect" based on who did the experiment.
@@statquest makes sense thank u! Yes it was same material in my case since all I'm doing is making the reaction mixture and pipetting same volume of that reaction mixture into 3 wells. So it is a technical replicate!
@@chw512 Note: Even if it is you doing the experiment on both days, there can still be a batch effect (maybe you sleep better before one batch of experiments etc.)
how do you combine technical and biological replicate variance if both are of the same order.....or if tech replicate variance is large and could affect the comparisons in the biloical replicates the object if the study
You have to create a complicated design matrix to accommodate both types of replicates. To learn more, see my series of videos on linear regression: th-cam.com/play/PLblh5JKOoLUIzaEkCLIUxQFjPIlapw8nU.html
Hi Josh,
Quick Question:
For cell lines, if I were to use only one particular cell line and repeat it.
If I have done this experiment 3 times, as compared to having 3 different measurements in 1 experiment.
Would that be considered biological replicates for repeating the experiment 3 times?
May I help you with this? Look, you are looking for the logic that you repeat a measurement. If you are going to do a cell experiment single cell line, you use technical replicates, which is the repetition in one experiment or repeating the experiment three times. All you are doing is that you measure something in a cell line. in the first, you measure it in three collection of the cells from the same cell line, in the latter you measure it along with the variation in different set of experiments. If you do the same experiment in multiple cell line that will be your biological replicates. Technical replicate is related to the techinque profile such as precision, data variation in the technique, etc.
Hi StatQuest, thanks for the video. It confuses me a bit because in the examples you give of technical replication, seems like an example of pseudoreplication, due to the samples are not independent of each other. Am I right? if not, why? Thank you again.
In this case pseudoreplication is the same thing as a technical replicates since the measurements are correlated. However, the goal in this case (for technical replicates) is to measure the technical error, and we do that by removing the biological variation - thus, we do the same experiment from the same biological sample.
@@statquest Great! seems clear to me now
Hi Josh. Thanks for your video. I am still confused about my case. Recently, I am collecting sample for RNAseq. My sample is pig oocyte. Each week, I can get 10 ovaries from 5 individual pigs (one pig contains two ovaries). Normally, one ovary contains 10-20 healthy oocytes so I can get ~150 oocytes I want each week. For establishing the RNA library, 30 oocytes is required. If I want to have 4 biological repeats (4 repeats need 120 oocytes), can I finish it within one week? Or I should collect 30 oocytes each week and finish sampling 4 weeks later?
You can get them all in one week.
Thanks
bam!
You are my hero❤
Hooray! :)
Anyone knows a book discussing about these statistics in biology? Thanks in advance
Hopefully I can include this in my book on statistics.
The replicates are like the GRR in the manufacturing industrial.
noted
If sample is coming from different strains of same bacteria in similar conditions, are these biological replicates?
It depends. If you have 3 different strains, then you have 3 different biological replicates.
@@statquest I don't understand. If I have a bacteria, and I have RNASeq data of 3 of it's strains under same condition, are these biological replicates?
You are awesome!
Thanks!
lol bro love the explanation and the cheesy intro music
Best!!
Thanks!
BAM! Thanks dude (:
Any time!
🙂🙂💪
Nice! :)
mah nigga
PS do not measure the orange dude. Please we do not need more of them....
:)