Python for Bioinformatics - Drug Discovery Using Machine Learning and Data Analysis

Learn how to use Python and machine learning to build a bioinformatics project for drug discovery.
✏️ Course developed by Chanin Nantasenamat (aka Data Professor). Check out his KZbin
channels for more bioinformatics and data science tutorials:
/ dataprofessor and / codingprofessor
🔗 And Medium blog posts for more data science tutorials / data-professor
⭐️ Code ⭐️
💻 Parts 1-5 github.com/dat...
💻 Part 6 github.com/dat...
⭐️ Course Contents ⭐️
⌨️ (0:00) Introduction
⌨️ (4:29) Part 1 - Data collection
⌨️ (26:57) Part 2 - Exploratory data analysis
⌨️ (49:41) Part 3 - Descriptor calculation
⌨️ (1:01:51) Part 4 - Model building
⌨️ (1:10:41) Part 5 - Model comparison
⌨️ (1:18:15) Part 6 - Model deployment
🎉 Thanks to our Champion and Sponsor supporters:
👾 Wong Voon jinq
👾 hexploitation
👾 Katia Moran
👾 BlckPhantom
👾 Nick Raker
👾 Otis Morgan
👾 DeezMaster
👾 Treehouse
--
Learn to code for free and get a developer job: www.freecodeca...
Read hundreds of articles on programming: freecodecamp.o...

[@Alecor_studio]

Never in my life would I think I would get a video for this.. There's not many resources on bioinformatics on youtube. This is godsend! Thank you very much.

[@DataProfessor]

Thanks for watching! Appreciate the kind words!

[@burntchickennugget191]

Ikr its actually really nice. My college doesnt like letting people take courses on codeing in general and are very picky with who can teach it. This is what I wish I had back 10 years ago.

[@karthikraja1029]

YEAH IAM TOO

[@codelucky]

Can a Machine Learning Engineer(CSE) get into Bioinformatics who did not have Biology at Bachelor's? Can it be consumed via self-learning?

[@esmael.c2b]

@@codelucky if you learn in a school, that's learning, if you self-learn, that's also learning.

[@clodsire_]

wish this vid came out sooner. i’m about to graduate as a bioinformatics major and i remember struggling so much bc of the lack of online resources and material. whereas you can learn anything online as a swe, you’re completely dependent on school for bioinformatics. to anyone wanting to pursue bioinformatics, it’s a very complex major but if ur passionate about bio, data analysis, and tech then it’s a great choice

[@DataProfessor]

Yes, bioinformatics is definitely an exciting field for applying SWE, ML and AI to help make sense of the big biological datasets.

[@cj1871]

What would it take for bioinformatics research to become decentralised? Maybe a community funded lab where remote researchers pay to have their experiments run and the results are public which allows other researchers to contribute?

[@AJ-qt9hx]

I'm just starting on Bioinformatics after my degree course in Biology. Thank you for putting this video together but this is no beginner course. I'll work on taking some professional course on edX and Coursera, then revert back to this. Thank you sir. ✨👏

[@louzewdie9104]

As someone with a Bioinformatics Minor who always wants to learn more, I cannot express how ecstatic and grateful I am to find your content.- Thank you good sir!!

[@helloworld2054]

I want to study bioinformatics and I'm currently learning python. So grateful for this course!

[@DataProfessor]

Awesome, glad to hear that!

[@tilbintt3491]

I have learned Bioinformatics. That was very hard tbh 😰

[@helloworld2054]

@@tilbintt3491 oh :/ what do you have to study there?

[@tilbintt3491]

@@helloworld2054 I had it in my masters degree. It was more related to Genes and stuff. Analysing gene data was a nightmare for me. I don't remember exactly but different set of genes or data had a different way of approach.

[@helloworld2054]

@@tilbintt3491 I see. So what are you doing now?

[@donutsandpenguins2839]

Can't believe how lucky I am! I was just considering my options among the hundreds of computer science branches yesterday, and bioinformatics and computational biology was the most interesting one I found. Since it's still new compared to the other branches, there wasn't much I could find out about it. Thanks a lot for the work put into this video!

[@DataProfessor]

Glad to hear that! Bioinformatics is certainly an exciting field!

[@punkkauz]

@@DataProfessor I wanna know more about bioinformatics, scholarship this and that... would you give some hints

[@vidyashreeshinde654]

You are on the right path

[@Dankman3]

Im starting a job at Astrazeneca next month. Thank you sir for the course, I will use the knowledge to the best of my ability.

[@mr.anonymous6640]

Congrats bro , Wt is ur qualification😅

[@Dankman3]

@@mr.anonymous6640 BE Computer Science

[@innokirigiti]

Hello, I would also like to know your current qualifications. I am interested in this field

[@vasudevkarthik4593]

A tab bit late . But I would like to congratulate you. Hope this ramdan you will get all happiness and good things that you have cherished in life. Peace

[@barisbasar3909]

I hope you are nailing it!

[@emmanuelonah4596]

Thank you professor. It's so heartwarming to have this quality of information shared for free. The best I have seen on Bioinformatics. All my questions answered and opened my eye to the many facets of computational drug modeling and simulation. Thank you once again, Prof.

[@miltonkambarami8129]

I am from the Virology side of Bioinformatics and its wonderful to find a Bioinformatics-related tutorial on a computational-based learning platform. I feel Bioinformatics/Computational Biology is starting to embrace more of the Computational side to answer Biological questions and the two are starting to merge as one. Sooner or later Programming is going to be a compulsory skill just like writing but what will matter is types of questions and problems you can solve using programming.

[@silvereyes000]

Virology side of bioinformatics? What is it about?

[@ai-ur5uv]

@@silvereyes000 probably anlysing viral genomes

[@jaggyjut]

This is Pure Gold. Thank you for creating this tutorial.

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@bro...5849]

@@jimcrypto8573 no

[@joyrainbowdress]

I'm a dentistry student but I love computers as much as I love biology, I also would love to explore the world of clinical research and pharmacovigilance. I think I'd take up bioinformatics in my masters along with my clinical practice, it seems like the best intersection between healthcare and IT.

[@ai-ur5uv]

OMG same, I do dentistry and thinking about getting into informatics but not sure because it is very broad and dentitry is usually clinical so very different skill sets

[@joyrainbowdress]

@@ai-ur5uv Actually I left my dentistry degree to pursue a bachelors in CS soon after I made that comment, my love for CS surpassed my interest in healthcare. But I hope you can break into informatics!

[@wintur2856]

@@joyrainbowdress I'm getting my Computer Science degree aswell. Afterwards, I'll be pursuing a Masters degree in computational biology or bioinformatics. and a Doctorates in Bioinformatics.

[@joyrainbowdress]

@@wintur2856 That sounds like such an amazing and well-thought out pathway, I wish you all the best!

[@shakuntalabaichoo7922]

This is the best bioinformatics video/tutorial on youtube, I have ever come across. Thank you so much for such a great tutorial/resource!

[@brunobustos96]

Computational neuroscience next please! Love your channel

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@BayAreaPressurePros]

The gethub items are not coming up for me, would you mind maybe uploading them? Other than that, these videos look amazing!!!!! my thesis is bioinformatics heavy and our bioinformatics course on campus is more like an introduction lol. thanks my friend, Cheers.

[@palakgupta8728]

its a great video really helpful. at 14:09 you are not able to see other types of standard type variable because youve already entered its filter as ic 50 so it shows only ic50 entries. once you remove filter.standard type[ic50] , youll see array with inhibition, ki, ec 50, kd, activity.

[@0307ismail]

Hi Data Professor, I am doing PhD in Bioinformatics and my research topic is the same as of this video. Thnx alot for this video. I think i will need more of your support and guideline for my research.

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@pacifio]

man be expert in biology and computer science while me just being dumb 24/7 epic style

[@leonardomontes2975]

You are always on top guys! I really appreciate the way you share the knowledge! Thanks Quincy!

[@bassamtork7679]

Huge work is put in this video, wonderful method of delivering the course. Many thanks to Dr. Chanin Nantasenamat

[@MrViperHiggins]

Love it man keep up the great work. Really appreciate people like yourself making data science accessible to those of us coming through the process of learning and working with data.

[@DataProfessor]

Awesome, thanks for watching!

[@azadjain8534]

I was searching for this content for long time. Finally this arrived. Thank you so much for invaluable content.

[@DataProfessor]

Awesome, thanks for watching Azad!

[@malasharma5512]

Mee too

[@jannmikoingelrabagogamingc6012]

Oh hey! I am so grateful and excited for this one man! I am an upcoming 3rd year BS Pharmacy student and planning to do my undergraduate thesis on this area specifically.

[@stringpriest]

Thanks so much for this video. I obtained an Msc in pharmacognosy and taking another masters in Regulatory science. Trying to go into tech, I started a course in Data analysis but the goal is Data science. I am passionate about bridging tech and healthcare as related to data science in my thesis which leads me here. Thanks a lot once again.

[@user-fe2oh8oj2u]

Looking forward for "Python for Finance"! :)

[@donaldderrick1595]

Algorithmic Trading for finance. on freecodecamp bro

[@user-fe2oh8oj2u]

@@donaldderrick1595 , thank you, I will check it out.

[@RohitSingh-yo2yl]

ML for finance

[@BerkshireHathawayCRE]

I hope you guys can make more bioinformatics videos! Such an important application!

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@benjamintwumasi2480]

Thanks so much for this video, Professor. You have saved a million lives of people who want to learn Python for drug discovery. Thank you once again.

[@e.m.2655]

Been hoping you would make this video! I have a background in biology and a very small bit of experience in informatics.

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@Teflon2000]

I've study industrial chemistry in college...this video is perfect

[@vpundir3024]

It’s very useful for doctors who are learning coding 😼

[@khanmubeen]

Respected Professor! I'm quite amazed of how you stepped out to share and create such a knowledgeable course for all of us. I love that. 🤗

[@paulynamagana7560]

I'm doing my PhD in Pharmacy and I wish I had seen your videos before... I have discovered my love for Bioinformatics, too late to change my PhD tho

[@immcguyver07]

Not all of us even have a PhD.

[@ccuny1]

Fantastic. Incredible material on this channel and this is no exception. Thank you.

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@ccuny1]

@@jimcrypto8573 This looks like a scam. freecodecamp, it's your logo but it doesn't look right. You might want to clean this up.

[@ayeshaafzal4884]

Loved the whole video. Recommendation for Part 4 in colab, to prevent error. In the last scatterplot building chunk use: ax = sns.regplot(x=Y_test, y=Y_pred, scatter_kws={'alpha':0.4}) instead of ax = sns.regplot(Y_test, Y_pred, scatter_kws={'alpha':0.4})

[@alreadyghosts7727]

STEELY EYED ROCKET MAN GENIUS GOOD HUMAN!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! That totally worked!

[@georgevoknerech228]

Thank you so much. I've never dreamed that we would've get a lesson on bioinformatics.

[@knockknockyoo5812]

I truly appreciate it for sharing this video for free.

[@jondoe8658]

thank you and to all the people who share their knowledge for free

[@DataProfessor]

Thanks for tuning in!

[@febaelsamathew9348]

This was a topic I'm searching for several months. Since I'm from a pharma background and started learning python, but don't have an idea how this will work for job applications to the Pharmaceutical industry.

[@TheSsTrike]

Congratulations Data Professor!!!

[@DataProfessor]

Thanks Linas!

[@andrewchen2349]

So good to see Data Professor on FCC again! Thank you!

[@DataProfessor]

Thanks Andrew for the support!

[@gurudeebanselvaraj8888]

Excellent session. Hats off to you @Data Professor. I really enjoyed the whole lecture. It is quite informative for budding researchers in Bioinformatics.

[@scign]

In the deployed app, you should highlight any significant differences in the removed columns since those parts of the fingerprint could be important for pIC50 prediction. This would clearly identify any targets that fall outside the distribution of the training set so the model prediction may be inaccurate. Also, what was the purpose of calculating the Lipinsky descriptors if only the PubChem fingerprint was going to be used in the model?

[@nikitamaurya4518]

Thank you so much for this project tutorial and for explaining everything in detail. I have completed part 1 and am excited to do all the other parts.

[@dojaibi275]

Am currently study Msc Bioinformatics. Thank you for this package.

[@DataProfessor]

Awesome, thanks for tuning in!

[@blessingama3606]

Good afternoon Doja

[@blessingama3606]

Please can you share your experience in your masters with me. Thank you

[@prajyotprabhu827]

Excellent!! Much appreciated efforts you folks are putting in. Thank you very much 🙏👍

[@mahmoudal-bassam2507]

This is very helpful! Thank you so much. I tried something like this to compare distribution of normal and -log10 data (import matplotlib.pyplot as plt fig, ax = plt.subplots(ncols=1, nrows=2) df_combined.standard_value.plot.hist(bins=50, ax=ax[0]) df_combined.standard_value.apply(lambda x : -np.log(x)).plot.hist(bins = 50, ax = ax[1]) plt.tight_layout())

[@aashishkatyal]

Dear Professor, It is always awesome to watch and learn from your sessions. Can you make some video regarding docking by using python?

[@shenglinjing7350]

The latest CDD_ML_Part_1_Bioactivity_Data_Concised.ipynb for part I generates some wrong data for me (NaN values in bioactivity_class, negative LogP values etc.), which gives totally different statistics in Part II if I use that data. However, CDD_ML_Part_1_bioactivity_data.ipynb works fine, and I get the same results as in the video.

[@kevindelgado2982]

Glad this information exist on KZbin, in my thesis I'll be using the Megamolbart model from NVIDIA :)

[@user-yb6hm1jz6s]

Thank you very much Data Professor for sharing the great learning resource!

[@VyshnavieRSarma-rb7ur]

Wow Thank you Professor... Really enjoy learning a lot from your videos.

[@DataProfessor]

Glad to hear that, thank you Vyshnavie!

[@lo8885]

Beginning Part 4 , big clap for all of us Hope to find the thesis subject

[@DavidRamirezdmramirezs]

Wow great video tutorial. I will try to implement it in my lab. Thanks!!!

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@WelingtonSilvaMusica]

That's exactly what I was searching for, what rich material! Thank u so much!

[@aleksandonov8413]

Back in the day I was reading Perl in bioinformatics and it was mostly regular expressions. I guess these days it is Python and machine learning.

[@GCKteamKrispy]

Just found a book about this and got excited about this field. It sounds interesting

[@shivanipawar2296]

Thank you so much sir for this useful information. it's very difficult to get any information about Bioinformatics on KZbin. I'm currently studying in BSC( Bioinformatics). Looking forward for more information.

[@crismo7753]

Thank You very much for this great tutorial! Sometimes I am asking myself how much do You wish to have a better equipment, but I don't dare to ask You that. I wish You great success and good health!

[@dumbkiddo3189]

This channel should receive an academic award in pedagogy. What you guys are doing is amazing!

[@belkhiranesrine185]

great work thank you sir, please can you told me why you choosed 'IC50" about which any information you selected it? Thanks in advance.

[@curiousresearcher308]

I guess IC50 (Inhibitory Concentration) tells us the information about the drug potential that how much is required for inhibiting 50% of the biological process for which the drug is being designed!

[@gustavojuantorena]

Yeah! Chanin has a great channel!

[@DataProfessor]

Thanks Gustavo!

[@johnraviella6561]

Please be advised: We may be able to model using an in vitro dataset, but we can only be inferential insofar as we can given the validity of our premises regarding these in vitro data. Basically what I'm trying to say is that there are unknown mechanisms driving favourable IC50s. Namely, Phospholipidosis has thwarted our computational repurposing efforts against SARS-CoV-2. Thank you for the video all the same.

[@DrNoureddinSadawi]

This is a fantastic tutorial, many thanks!

[@dangerday4277]

What timing. I was learning about it

[@OliverShey]

HI Sir, Thanks so much for the wonderful presentation. I have never watched such a clear video on the subject. Keep it up. I have another concern. I changed the target protein and could not get the dataframe displayed as with the Coronavirus example. I can see more than 27 targets when I search in Chembl. Is there another command or data upload process I should employ? How do I go about this?

[@deeptibhanot762]

Hello Professor, I want to work on parameter tuning of 3D Bioprinters before they print, so it will take input as the tissue type /bone type that we want to print and finally I want to predict the best possible combination of printer parameters for the job. Please help me from where I can get dataset for 3D bioprinted organs as well as dataset of 3D bioprinters.

[@devpatel2309]

why are you just running the code instead of explaining ?

[@nattawra]

Thank you so much for your wonderful video presentation, this is really helpful !

[@jaggyjut]

Would be able to do a session on HIPAA complaint application design. It will be interesting to know how to design the backend\database.

[@valentinaortiz9935]

So, correct me if I am wrong, but what you just taught us is how to potentially find a hit through a high-throughput computational screening?

[@anonymous-tt2lm]

Bio+informatics = new world

[@SUNYBOI]

I appreciate your effort

[@hanstsahnang4533]

Thanks soo much for this awesome tutorial. Can you work with an unknown library like for example you isolated a series of molecules in the lab and you decided to build a chemical library composed of those molecules. How to you load them on your working space.

[@miracleuche586]

Thank you so much for this amazing tutorial.

[@marcosdantasguimaraesfilho7747]

THANK YOU SO MUCH FREE CODE CAMP!!

[@nikitamaurya4518]

Part 2: I made a scatter pott, IC50 data is uniform but pIC50 is not. Do I still have to normalise the IC50? Also, the standard value of IC50 is not more than 1 million. In this KZbin video, you have used bioactivity_data_preprocessed file for part 2 and in the GitHub part 2 link you have used bioactivity_data_curated file.

[@hariks6880]

we have studied these in 2014 as Bachelors in bioinformatics (engineering)

[@rachelk2600]

Please release an updated AI bioinformatics course!

[@ragnarok3888]

Pls don't forget about the ursina engine, we need a video on it

[@Anna-zo2tv]

you are angels for creating this ❤

[@vishnus5531]

Thanks for this course.... 😍

[@christianscientist3963]

Since you used the CHEMBL database to look at targets and small molecules in your course, is there a library code to extract Zinc and PubChem drug database to use their small molecule?

[@zahrayzd2842]

Thank you for.this is great 😃👍

[@science_mbg]

Because you did not get any other stardart_type is that you filtered to IC50 only one line before so that is it only unique for IC50.

[@militant_dilettante]

Thank You for such wonderful lesson! So much useful information, and so tightly packed! I have questions, though. In the beginning of both notebooks for Exploratory Data Analysis we go through the steps to modify the "canonical_smiles" variable, namely, we seem to remove parts of SMILES strings that denote smaller molecules not connected by the covalent bonds (not my area of expertise, not my first language, therefore such an awkward phrasing), and leave in place only the biggest part, probably because it is much more likely to be biologically relevant. Is this step specifically for calculation of the Lipinski's descriptors? Because later we concatenate the descriptors dataframe with the original dataframe. Also, I am doing that lesson more than 2 years after the video was posted, and in the Acetylcholinesterase data ChemBL fed me at least one "standard_value" value of 0. What do You recommend replacing it with: 1 or NaN?

[@kuldeepkaur-do2ex]

How we can retrieve fda drugs data directly from zinc database as well as the biological activity of those drugs? Thank you

[@theaieducator1595]

Bioinformatics is interesting...Thanks for the helpful videos

[@indumatisharma3646]

Thanks for this video. Can the model be deployed in RShiny?

[@jagadeeshpoyya9699]

How to validate QSAR Model and how to utilize finger print to generate the Williams plot

[@mangalahegde3805]

Thank you Professor. I have a doubt. You explained with a example of a target. How can we do the same with multiple targets like all the members of one protein family as targets?

[@apoorvasetu5840]

Hello Chanin, I am using EC50 standard_type for Acetylcholineesterase. When labeling compounds, is it the same as IC50 or is it different?

[@tahircak1375]

Thank you Professor

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@nayeemanushrat3174]

Hello Data Professor, first of all, thank you very much for this awesome video but I'm facing some issues with the EDA part! Could not install the "rdkit" ! so can't move on to the next part of the coding..can you please help me with this, I really want to learn it, thank you, looking forward to your kind response ☺

[@aryahi9220]

try adding "!pip install rdkit" at the beginning of the code of importing libraries. that worked for me, and it might work for you as well!

[@iii1090]

Hello,Professor,Thank you for your hardworking,could you please inform us whether we can open the same.ipynb as your,or we need to input the code in python or other software.as I am a beginer,so please😄

[@tejiyo]

In part 3, Why did you convert the df to .smi file and used command line to apply padel function, can you explain the line " ! bash padel.sh" please

[@danys9085]

Thank you for this amazing video but I wish you could be a bit more specific , I feel sometimes lost

[@vasudevkarthik4593]

Python can be used for biometrics as well. That is surprising. I thought python would lose its value in 2022 but it appears pytjon will stay for another 10 years

[@epicrj]

Thank you 💖

[@nicolastarquino1061]

Hi Master, thank you for this great video!!!!!!!!! in the minute 46:59 there is a cut that not allow to understand the idea

[@duartelucas8129]

this is really cool

[@jimcrypto8573]

Thank you for watching, for question, guidance and support at anytime. Just DM Me +1.(.9.1..4). 7..0.4..4.3..3.8..@ W H A T S A P....

[@sarahchenaa786]

I'm biologist & intrested Please let me know if I can start this course as a full beginner in programming ?^^

[@sarahchenaa786]

@@mexicanburrito138 thnx ^^

[@thinamG]

Awesome!

[@DataProfessor]

Thanks Thinam, glad to see you here!