Mike Saint-Antoine

Пікірлер

@castilloh.gianmarco1048 Күн бұрын

gold chains!! <3

@summergeary8581 Күн бұрын

Hello mike, i love your videos I haven’t found anything els like them! If possible , could you please one day do a tutorial on how to model how cells interact via Bioelectric signals on python. Would be super cool! Once again thank you for your videos and being a spectacular teacher :)

@rsbenari Күн бұрын

Crystal. And completely helpful. Thanks so much!

@williamament7826 3 күн бұрын

No way, I subbed to you a while ago when I was getting looking at career paths as a Computer Science & Data Science graduate. I've been exploring Julia in my free time because I also want to get into parallelizing my programs. The multiple dispatch feature of Julia is super interesting to me. Keep making Julia videos!!!! <3

@MikeSaintAntoine 3 күн бұрын

Oh nice! Yeah I had to learn Julia for my job and it took a bit of getting used to at first, because I was so used to Python. But I'm actually liking it now! And yeah there will definitely be many more of these videos coming 🙂

@ada8189 6 күн бұрын

Apt. Thank you

@shv_ 7 күн бұрын

What do you think is the best background to get into bioinformatics? I’m thinking to do CS as Bachelor and then Bioinformatics as master. Do you think studying CS as Bachelor than Biology or Data Science? Thanks in advance !!!

@benklassen77 16 күн бұрын

This is very helpful for my research. How would you change the loglikelihood code if you wanted to account for the prior probabilities, noting that P(theta|D) ∝ P(D|theta)P(theta) rather than just P(theta|D) ∝ P(D|theta)?

@chrislee8408 19 күн бұрын

Is there a way to still normalize the counts data if you only have the final DESEQ results table saved as an excel/csv file? Such that, you don't have access to the original dds object anymore..?

@mikaeelahmad7594 21 күн бұрын

God bless you

@triwidjajanti2437 25 күн бұрын

Hi Mike, how to draw the phase portrait for predator prey system with three equations with phyton? Thankyou.

@MikeSaintAntoine 25 күн бұрын

Hi! I've actually never done this myself, but here's an explanation I found of how to make 3D plots in Matplotlib, so I think this is the best approach to plot the phase portrait for all 3 variables in the system. jakevdp.github.io/PythonDataScienceHandbook/04.12-three-dimensional-plotting.html If you run into any trouble you can always email me your code at [email protected] and I can try to help out with it. Thanks for watching! 🙂

@triwidjajanti2437 25 күн бұрын

Thankyou Mike.

@DavidValle-ej8es Ай бұрын

When you said reverse dogma I thought of the HIV mechanism!

@MikeSaintAntoine 25 күн бұрын

Yes, that's exactly what I had in mind! The HIV mechanism is the opposite direction, converting viral RNA to DNA with reverse transcriptase!

@Purnima_Chowdhury Ай бұрын

is there a way to reduce the font size in the graph

@MikeSaintAntoine 25 күн бұрын

Hi Purnima, sorry about the late response! I think you should be able to adjust the font by changing fontsize_row and fontsize_col in the pheatmap() command. I think there's also another option to just set a general fontsize in this command. Anyway if you're still having trouble you can always email me your code at [email protected] and I can take a look and try to help. Thanks for watching! 🙂

@federicosisti2608 Ай бұрын

Short and helpful. Is random a sign of you don't know some variables? If I add new variables can I transform a model from stochastic to determinist?

@MikeSaintAntoine 25 күн бұрын

Hi Federico! Hmm that's an interesting question. I think that's true from a philosophical standpoint, in the sense that a lot of the things we model as "random" aren't truly random and are actually deterministic, but they're being determined by deeper, more complex things that we don't have enough information about so we just treat them as if they're random. Like in the gene expression bursting example, we model the bursts of gene expression as occurring randomly, but if we were to make the model more complex and track the exact position of every transcription factor floating around in the cytoplasm then the bursts of gene expression would NOT be random, but would happen when the TFs bind to the DNA. Another example -- technically we could say that the result of a coin flip isn't random, because it's determined by the starting position of the coin and the laws of physics. But tracking all of those things to model a coin flip as a deterministic process would be insanely hard, so we just treat it as random since we don't have enough information to predict the outcome either way. So I think that's an interesting philosophical point. However in practice, the goal of our models is not to be perfect, but to be useful. So a lot of the time, treating these things as random is useful for us when we don't have enough information to track the true deterministic causes of them, or we just don't want that precise a level of detail when approximating the thing as a random variable will do. So sometimes it might make sense to try to convert a stochastic model to deterministic by tracking more variables, but other times it might be better to just keep the model simple by incorporating randomness. Thanks for watching! 🙂

@livelife8130 Ай бұрын

You discuss it well, thank you so much.. from someone who doesn't have formal education on bioinformatics and is also a self learner

@MikeSaintAntoine 25 күн бұрын

Thanks for watching, and let me know if you have any questions! 🙂

@sapienthought1103 Ай бұрын

hello thank you for this video could you please make a video on how to download raw fastq files from GEO ?

@MikeSaintAntoine 25 күн бұрын

Yeah that's a good idea, I'll try to make a video on that in the future!

@idrisbabaji-tm9ln Ай бұрын

Lovely

@MikeSaintAntoine Ай бұрын

Thanks for watching! 🙂

@zebmason6530 Ай бұрын

Hadn't come across the Gillespie algorithm before. Obviously produces the wrong distribution for epidemics, kzbin.info/www/bejne/fWWvmIagfNRpjsU, wonder whether it is correct in its other applications.

@arjunshah7105 Ай бұрын

What's the answer lol

@MikeSaintAntoine Ай бұрын

Hey Arjun! The answer is that the points at the end are NOT steady states. One way to confirm this is to try plugging these points in to the ODE equations. In order for a point to be a steady state BOTH ODE equations need to equal zero, or else some change will be happening in the system. But if we plug these points in, we see that they cause one equation to be 0 and the other to be some number. So that's the mathematical way to check for steady states, but we can also reason about it logically to reach the same conclusion. For the first point, there is a positive number of the prey, and 0 predators. What would happen if this were the case? Well without any predators to kill the prey, the prey population would just keep increasing -- this means that change is happening in the system, so it isn't a steady state. For the second point, there are 0 prey and a positive number of predators. What would happen if this was the case? Well, without anything to eat, the predators would gradually die off -- this is a change happening in the system, which means that it isn't a steady state. Thanks for watching! 🙂

@negarm9006 Ай бұрын

Thanks a lot! This is exactly what I was looking for!🙏

@MikeSaintAntoine Ай бұрын

Thanks for watching! 🙂

@Parallaxxx28 2 ай бұрын

Hello Mike! I stumbled upon your channel when searching for the difference between stochastic and deterministic models and I find it really interesting how to apply data science and mathematical concepts in bioinformatics. Can you recommend some books that cover the math that you teach in these videos? I want to read into more theoretical math. Thank you and I hope your channel grows, it really is a treasure-hove of so many real life applications of data science and ml!

@MikeSaintAntoine Ай бұрын

Sure! I'm not a big textbook person and usually prefer learning from tutorial videos, but if you're looking for a textbook on computational biology and math modeling then my favorite one is Uri Alon's systems biology textbook: amzn.to/4a6WIlM For machine learning, if you're a beginner then I'd recommend Sentdex's textbook on how to make a neural network from scratch in Python: nnfs.io/ But if you want something on machine learning that's more advanced, I'd recommend this one: amzn.to/4duhXAQ I hope that helps! Thanks for watching 🙂

@Parallaxxx28 Ай бұрын

@@MikeSaintAntoine Thanks for the resources Mike! Sorry for the late reply 😅

@sajanphutela4818 2 ай бұрын

Thanks mike for such nice video on MCMC. This is first time, I could understand this. What is you email address? I I have some questions if you could help me with that?

@MikeSaintAntoine Ай бұрын

Thanks for watching Sajan! And sorry about the late response but yes I got your email and replied to it 🙂

@sheetal_soul 2 ай бұрын

Hello .. thank you for this video. I would like to know how to see 'Correlation analysis between phenotype and transcriptome'. Can you please let me know how to do this?

@MikeSaintAntoine 2 ай бұрын

Hey Sheetal, I saw your email and responded with a couple ideas! 🙂

@alemneshhailemariam2541 2 ай бұрын

Thanks

@MikeSaintAntoine Ай бұрын

Thanks for watching! 🙂

@KealyGlenn 2 ай бұрын

what do the alpha, beta, and other variables represent. Would it be the growth rate and death rate of each animal?

@MikeSaintAntoine Ай бұрын

Hey Kealy, sorry about the late response but yes that's correct!

@livialopes5682 2 ай бұрын

very helpful! Thanks :)

@MikeSaintAntoine Ай бұрын

Thanks for watching! 🙂

@infinger2006 2 ай бұрын

Thank you - clear, concise, and to the point. I think I could use a stochastic model idea for manufacturing. We have tons of randomness in the process, everything from materials to operator variation, to when things are pulled from a freezer before use.

@MikeSaintAntoine Ай бұрын

Thanks for watching! And yeah I think there's an entire field of stochastic modeling devoted specifically to industrial applications like manufacturing, and how to make things as efficient as possible when operating under uncertain conditions. Interesting stuff!

@shayanakhavan6002 2 ай бұрын

Wow this is amazing, Mike! I am about to defend my thesis next month for my bioinformatics masters, so this will be super helpful when I am searching for a job!

@MikeSaintAntoine 2 ай бұрын

Oh nice, congrats! And good luck with your job search, there are a lot of exciting opportunities out there.

@aseemchib2233 2 ай бұрын

Why isnt there a training or test split in this model?

@MikeSaintAntoine 2 ай бұрын

Great question! Usually training / testing splits are used in supervised models, where the goal is to train the model to recognize a pattern on a set of data where the correct answer is given to it, and then test it on new samples that it hasn't seen yet -- actually this would probably apply to the traditional KNN algorithm when used for clustering/classification. But in this video, we were taking an unsupervised approach and simply trying to identify which points were the furthest away from the rest of the points, rather than trying to classify or cluster the points in general. Anyway I hope that helps. Thanks for watching and let me know if you have any more questions! 🙂

@reasons4171 2 ай бұрын

Awesome!

@MikeSaintAntoine 2 ай бұрын

Thanks! 🙂

@shiyiyin3403 2 ай бұрын

great work!

@MikeSaintAntoine 2 ай бұрын

Thanks! 🙂

@MikeSaintAntoine 2 ай бұрын

Hi everyone, thanks for watching! Here's all the code from this series I'm doing on transcriptomics in R: github.com/mikesaint-antoine/Comp_Bio_Tutorials/tree/main/transcriptomics_with_r

@MikeSaintAntoine 2 ай бұрын

Hi everyone, thanks for watching! Here's all the code from this series I'm doing on transcriptomics in R: github.com/mikesaint-antoine/Comp_Bio_Tutorials/tree/main/transcriptomics_with_r

@MikeSaintAntoine 2 ай бұрын

Hi everyone, thanks for watching! Here's all the code from this series I'm doing on transcriptomics in R: github.com/mikesaint-antoine/Comp_Bio_Tutorials/tree/main/transcriptomics_with_r

@MikeSaintAntoine 2 ай бұрын

Hi everyone, thanks for watching! Here's the code from this video: github.com/mikesaint-antoine/Comp_Bio_Tutorials/blob/main/job_history_timeline/make_timeline.py

@lloydy272 2 ай бұрын

I just wanted to add that what you described as the Central Dogma of Molecular Biology but that is a mess up of what the Centra Dogma was supposed to mean. Crick coined the term Central Dogma of Molecular Biology and specifically stated that all forms of information movement between the different biopolymers was possible except for protein to nucleic acids. At the time RNA to DNA had not been shown but Crick stated that it was possible. Jim Watson then published the DNA to RNA to Protein model in a text but and wrongly called it the Central Dogma of Molecular Biology. That spread the misunderstanding of the interesting and specific point that Francis Crick was trying to make. What Watson was describing was just "how a gene is expressed" and either liked the term Central Dogma of Molecular Biology or misunderstood what Crick had meant and assumed overlap. Also Crick later admitted that he regretted the use of Dogma in the naming of this as he just liked the sound of it and did not realise that it was supposed to be unquestioned in its religious context, which is not compatible with science.

@MikeSaintAntoine 2 ай бұрын

Ah, I see! Good to know, thanks for the information 🙂

@lloydy272 2 ай бұрын

@@MikeSaintAntoine No problem - it is a common misconception that most academics get wrong. Also, thank you for the great video series. As someone who codes in python for genomics but is bad with maths, this was very useful for me to expand my understanding and capabilities.

@MikeSaintAntoine 2 ай бұрын

@@lloydy272 No problem and thanks for watching! 🙂

@kangqi8694 2 ай бұрын

❤

@IykeDx 2 ай бұрын

How did you get the lecturers email for cold emailing?

@MikeSaintAntoine 2 ай бұрын

A lot of the time they're listed on the university department's website. Like if they have a directory page that lists all the faculty, I'd recommend checking there.

@OghmaNano 2 ай бұрын

I really like this video and keep coming back to it. I think it's one of the most clear explanations of MCMC I've seen. You are a very good instructor,

@MikeSaintAntoine 2 ай бұрын

Thanks for the kind words! Yeah it's nice to know people are finding these videos helpful. Let me know if you have any questions 🙂

@CPATuttle 3 ай бұрын

I'm about to enlist soon and red-green color blind. Thank you for the video

@MikeSaintAntoine 2 ай бұрын

Good luck with your enlistment and let me know if you have any questions!

@CPATuttle Ай бұрын

@@MikeSaintAntoine I just passed the asvab and physical to get in. I did not pass the color blind test. One career not on here they said I qualify to join as color blind enlisted is the Intelligent Specialist. My leading two are storekeeper and intelligence specialist. Maybe that's a recent change because the panflit I received does say intelligence specialist cannot be color deficient. The meps recruiter liasion also said his opinion is that they will change the officer rule on color blind eventually.

@MikeSaintAntoine 25 күн бұрын

@@CPATuttle Oh nice, that's great! Yeah back when I was joining intelligence specialist required normal color vision.

@bcj8894 3 ай бұрын

Thank you for the video

@MikeSaintAntoine 2 ай бұрын

Thanks for watching and let me know if you have any questions! 🙂

@bcj8894 3 ай бұрын

I love ur video

@MikeSaintAntoine 2 ай бұрын

Thanks for watching! 🙂

@Mostafa_99952 3 ай бұрын

Hello Mike, I hope you are doing well. I always watch your KZbin videos and learn a lot. I just saw one of your videos where you used the MCMC method for estimating parameters in MATLAB. I just have some simple questions about this method. (1) Is this MCMC method for parameter estimation the same as the Bayesian method? If not, then in which part of the code we can modify to make this Bayesian? If you share the code, then I will be grateful. (2) If we want to apply MCMC to a model where we have two or three species, then do we still keep the "normal" distribution that you did for the very last portion of the code? For example, for a two-species model, do we modify it like this --- loglikelihood = sum(log(pdf('Normal',data(:,2:3),testdata(:,2:3),1000))) Does that mean we just need to change the dimensions of the data and testdata only? (3) How do we transition from "normal" to "gamma" or "uniform" for the previous line? Also, I saw that the final output graph for estimated parameters for "r" and "K" looked like a normal distribution graph. You used the histogram option to plot the estimated parameters. My question is: is it the case that you choose "Normal" inside the loglikelihood code, I mean, the above code that I indicate in my question (2)? If I want to plot it like a normal-shaped curve, how do I graph it without a histogram? I tried the command plot(paraset(1,:)) in Matlab, but I don't know what's my y axis here. (4) If you share the background theory and reading behind this code, then it will be great. I can read and clear up some of the questions that I am looking for. (5) Do you have the Matlab code for the stochastic version of the MCMC of this example? How do we do the stochastic version of it? Thanks so much!

@MikeSaintAntoine 2 ай бұрын

Hey Mostafa! Sorry about the late response on here, but I got your email and replied to it, so hopefully that will help answer some of your questions. Thanks for watching! 🙂

@simonvalerioneumeyer3766 3 ай бұрын

Thanks or the good explanations! What measure do you use for goodness of fit? Just regular R-squared?

@MikeSaintAntoine 2 ай бұрын

Hi Simon! The loss function I used is basically just taking the difference between every prediction and corresponding data point, squaring it (to get rid of negative numbers), and adding them together. So it's basically just a straightforward measurement of how far away the model predictions are from the data. So this is the measure for how well the model fits, although technically it's a "Badness of Fit" rather than "Goodness of Fit" measure. I don't really like to use all the jargon and technical terms because it can get really confusing, but this type of loss function is typically called least squares, sum of squared residuals, or sum of squared errors -- and these all basically mean the same thing. Importantly, this is NOT the same as the R^2 statistic, although they are closely related. R^2 is the proportion of variance in the data that can be explained by the model. So to find R^2, you take the sum of squared residuals (our loss function), and then you calculate the total sum of squared residuals relative to the mean, divide them, and subtract this from 1: R^2 = 1 - (SSR/SS_total) So then you're basically getting a measure of how much of the variance in the data is accounted for by the model. Thanks for watching and let me know if you have any more questions! 🙂

@laminarflow6072 3 ай бұрын

Have things changed since you last posted this video that maybe you're aware of?

@MikeSaintAntoine 3 ай бұрын

Hey! Not that I know of, although I haven't been following that closely. The only change is that I think they added a new rating related to talent recruitment, but I'm not sure if it's open to colorblind people or not.

@jessicathomas4537 2 ай бұрын

Thank you for this video! My son wants to be in the Coast Guard reserves but he is colorblind. He also wants to study environmental regulations. It sounds like he wants to do almost exactly your career path. You answered so many questions for us.

@MikeSaintAntoine 2 ай бұрын

@@jessicathomas4537 Oh nice, yeah sounds like going MST could be a good choice for your son! If he has any questions about it, he can feel free to reach out to me with an email at: [email protected].

@shahab4804 3 ай бұрын

Retrovirus is an example when information flows the other way in nature

@MikeSaintAntoine 3 ай бұрын

Yup, that's right! 👍

@shahab4804 3 ай бұрын

Retrovirus is an example when information flows the other way in nature

@neerajdhandia7876 3 ай бұрын

Reverse Transcription where reverse transcriptase enzyme make DNA from RNA, takes place in virus i.e. HIV to replicate their genomes.

@MikeSaintAntoine 3 ай бұрын

Yes -- that's right!

@reasons4171 3 ай бұрын

This is really cool! I’m excited to follow along during this weekend. I like how your tutorials are fairly beginner friendly and well explained. The topic is intriguing as well! Thanks for making these videos man

@MikeSaintAntoine 3 ай бұрын

Thanks for the kind words! Yeah it's nice to know that people are finding these videos helpful. Let me know if you have any questions 🙂

@vinicius.config 3 ай бұрын

Obrigado irmão! Será muito bom para meus estudos, estou tentando entender o modelo e trazê-lo para o código. Obrigado!

@MikeSaintAntoine 3 ай бұрын

Thanks for watching! 🙂

Ең жақсы KZbin

Пікірлер