The term pleiotropy means a single gene that has multiple functions; I chose that name for the channel because I have a wide range of interests, scientific and otherwise.
I'll be posting videos about evolutionary biology, statistics, and other (mostly scientific or quantitative) things I find interesting.
I'm also developing some full function websites linked below.
Feel free to send me a message and let me know what you've appreciated and what you would like to see more of.
For the logistic regression slide, are you sure you don't mean 1 + e^-z, not 1 - e^-z? Using subtraction in the denominator seems like it could cause a lot of issues.
Hi, thanks for this video. I understand that its a few years old. But, I was hoping you'd be able to provide a citation for the F-max table included in the video. In the table, it states where uneqaul treatment sizes, use the smaller df - would be grategul if you could provide guidance/citation on where this statement comes from, i.e., to use the smaller of the two dfs when unequal samples? Thanks.
I don't have an exact citation, but this statement is based on the general principle of assuming you have less data than you do so that you are biased towards making conservative conclusions (i.e., less likely to reject H0).
Amazing, thank you! Yes, it would be good to know the citation for that principle, a landmark paper or something that I can use to justify using the smaller of the groups/df for my analysis.
Biology degrees and also any job you get will pay you whatever you want. It just takes time and experience and the guts to tell people what you deserve. Also with more money comes more responsibility, but you get paid for the work you produce. Take pride in your work and have some self respect. Biology degrees are definitely worth it and if you look into getting your MLT or CLS it will pay upwards of 80-130k and you don’t need a grad MS degree just a bachelors and the MLT or CLS license
I do find it a little annoying to have to rely on RU-vid to find sufficient information about statistical functions - a job that Wikipedia should be doing infinitely better. The math/stats oriented sections on Wikipedia are somehow awful, unreadable, _and_ incomprehensible all at the same time. This video saved me a lot of pain. Thank you for putting it together. Unfortunately, there is only a tiny audience who needs the information that I was seeking and found in this video so it probably won't exactly go viral.
The output of the heteroscedastic t-calc formula is roughly "in the neighborhood" but not sufficiently accurate, only producing around 2 decimal digits of precision in Microsoft Excel. The homoscedatic and paired t-calc formulas in your videos produce identical results in Excel (15 digits of precision). So we can hypothesize that either the heteroscedastic formula in this video is wrong _or_ Microsoft Excel is wrong. Wikipedia uses similar notation to what is shown in this video (but far more obtuse - thanks Wikipedia!), so I'm inclined to believe and conclude that _the heteroscedastic formula in this video is correct_ and that _the Microsoft Excel T.TEST() function produces incorrect output_ for the heteroscedastic t-test.
I do plan to update the videos as I go, but covid is a nice example of one of the things I talk about in this video. Since covid kills very few people (as a percentage) and is asymptomatic in so many, that's a big part of why it spread so well. If it killed more and always had severe symptoms, then it would have spread more slowly.
This is a wonderful video, and super helpful for someone who is going to be going into a biology major soon, but my only complaint is the music, it was so loud that at times it was hard to hear your points. Other than that, thank you so much for the great video!
Glad you liked the video. As for the music, unfortunately RU-vid does an auto-adjust which can sometimes raise the background volume even when you make a video with soft background music.
The pdf powerpoint has all the info in one sheet which makes it really confusing to read. How can I see each power point individually? or maybe just convert it?
But isn't there a mutation type known as subsitution mutation? Wouldn't using the term for a different meaning complicate things? Nonetheless, i was able to understand polymorphism through your video. Thank you.
Can you discuss the effects of females’ mate selection while on hormonal birth control? Lately, I’ve seen some discussion about this and the possibility on issues since the hormones influence the females to choose mates with less typically masculine features. Thank you.
Is Sexual Selection for baby-face, selected itself by the advantage of having larger brain? I think the question can be answered by asking another question: Does peadomorphosis happen for all features of an organism at the same time, like if I am peadomorfic for some reason I should have all the features of a child, like voice, body proportions, etc.? This is my question....
All matter, living or otherwise, eventually ceases to exist. This is a consequence of entropy. The more time moves forward, the more everything becomes disordered.
It's true that entropy increases in the overall system, but localized parts can become more organized (at the expense of other parts). As an example, the Earth becomes more organized (e.g., life etc.) because the increase in entropy of the sun outweighs it in the system as a whole (i.e., the solar system).
The biggest frustration I had with statistics is that standard deviation is presented as the primary descriptive statistic for measuring spread. Standard deviation has useful mathematical properties for further calculations when doing inference. But for description, IMO the mean absolute deviation is just obviously superior. It's intuitive. You know what it means. And it makes sense even when the data distribution doesn't fit any known distribution (non-parametric). It's more robust, in the technical sense. The only answers I could ever get from stats professors about why they were doing this was (a) mean absolute deviation is not differentiable and (b) standard deviation tells you how far out to go on the distribution to cover x % of it. (a) is only a good reason when dealing with inference. (b) is only true when you know the distribution, but in those cases you can just use mean absolute deviation anyway because it also covers a fixed proportion of the distribution. When there is no distribution, standard deviation is impossible to interpret. Mean absolute deviation is not. What do you think?
Thank you for this. The part that I have difficulty understanding is that, mutations seem somewhat rare, a mutation that happens to be in a gamete seems like it will happen very seldom, a mutation that happens in a gamete, and produces an advantage, seems like it will almost never happen, and a mutation that happens in a gamete and produces an advantage and then that organism doesn't get killed just due to bad luck before it can reproduce, seems like it should happen almost never. Are there other sources of new genes other than mutation? Are there other "fuels" for evolution other than new genes? Or are mutations just a lot more common that I'm assuming? Thank you.
Mutations are the only "fuel" for evolution, but there are many different kinds (e.g., change of nucleotide, duplications of a length of DNA, etc.). They are extremely rare for any given location, but the genome is huge, populations are often quite large, and this has been going on for millions of years = lots of fuel. As you correctly note, most new mutations are lost due to random chances, but there are lots of mutations happening all the time.
@@pleiotropy Thank you for the response. The different kinds of mutations is interesting I'll look into that. I'm going to follow up with more questions but feel no obligation to reply if you busy. 1) The "millions of years" factor, my understanding is that the reason why most big changes take millions of years is not because evolution is slow but because the environment usually changes slowly, but that when there is a sudden change in an environment then many species can undergo a large number massive changes often in only a few dozen generations. So like, when trying to understand how evolution is able to move so quickly, I think it's a bit besides the point to say millions of years = lots of fuel since even in tens of years there appears to still be more than enough fuel if that fuel is actually needed. So I'm still thinking there must be a hell of a lot of mutations for that sort of thing to be possible? 2) Do we have any idea of quantity here? So like, for example, in lets say a dog: (a) what order of magnitude of mutations are you getting in the dog in its life overall and (b) what percentage of those are in the gametes and Like are we talking dozens? Hundreds? Millions? Billions? And what sort of tradeoff is there with cancer? Is it more fuel = more cancer always? Or can you have relatively large number of mutations without disproportionately large amount of cancer? Thank you again really appreciate it.