Imagine you’re conducting a survey in the school  cafeteria, asking questions and taking detailed   notes about how people like their pizzas.

We’re talking favorite toppings,   what kind of crust, and the ultimate  question: do you eat it folded or flat. Personally, I prefer it on a bagel,  cause then you can eat pizza anytime.

By the end of this survey, you’ll  probably have gotten some weird looks,   but at least you’ll have a big pile of data.  On its own, your notebook of pizza preferences  and tally marks isn’t gonna tell you much.  To make sense of those numbers,  you’ll have to analyze them. This could be as simple as counting up  your numbers and learning that more people   are rockin’ thin crust than deep dish. But what if you wanted to answer a more   complicated question, like how someone’s stress  levels affect how much or how little they eat?  Or, what if you wanted to take your survey  further and analyze the pizza habits of   the whole school district?

You’re gonna  need more than tally marks in a notebook. Biologists, and really most scientists,   use computers to analyze data all the time –  and for way more than planning pizza parties. Hi, I’m Dr.

Sammy, your friendly neighborhood  entomologist, and this is Crash Course Biology.  Hey Callie, could you serve me  up a pizza that theme music? (THEME MUSIC) Bioinformatics can help scientists sort  through data about everything from DNA   to weather conditions, to the number  of organisms on a beach three Tuesdays   ago — all information that can be useful for  different areas of research and conservation. You start with a question or problem. It can be as simple as pizza preferences,   or more complex, like, are two different species  of fish experiencing different amounts of stress?

Then, you get a collection  of numbers or a dataset.  It could be an enormous one showing, say,   the stress hormones of every species  of fish in the African great lakes.  Or, it could be a much smaller dataset about  the five jellies in your home aquarium. How ya feeling, buddies? Oh, they’re not…they’re not real.  We got ten episodes left and now y’all tell me.

Why have I been feeding them?! Ahem, anyway. So, the data could be   something you and your team personally collected,  or it could be from a larger, public dataset.  The bioinformatics community is pretty open;  they typically love to share information.

In fact, scientists are often  supported by data that ordinary   people have collected through community  science projects to help fill in gaps.  All that collected data can be analyzed  to help answer the original question. To find those answers, you take that data,  and you use a specific set of instructions   to analyze it, called an algorithm. In many cases, you communicate those   instructions to a computer, which  uses the algorithm to solve a problem.

You’ve probably run into your  fair share of algorithms around   the internet, like on TikTok or on YouTube. Like, if this video appeared on your YouTube   homepage, it wasn’t a random action. It’s because an algorithm analyzed   data about other videos you’ve watched and  predicted that you might like this one, too.

In this case, the “instructions”  are to learn your viewing habits,   and the “problem” is, well, how to  keep you there as long as it can. So be sure to smash that subscribe button! Sorry...we're obligated to say it  that way at least once per series.  Could you –uh–could you also  click the little bell too?

But algorithms can be fed all kinds of  instructions that help us learn some   pretty amazing information about living organisms,   the environment, or even the  patterns of disease spread. And bioinformatics can help us in nearly  every stage of the scientific process.  Like, let’s start with acquiring the data itself. For example, if we want to know the entire   genetic sequence of an organism, we can’t  just read it from one end to the other.  Because it’s so long, we can only  get it in small, overlapping chunks.

But we can then use computers to  organize those chunks and piece   them together in the correct order, like  assembling a jigsaw puzzle but in the end,   you get a complete dataset – and unlike a  puzzle, your cat can’t knock that off a table. Using that method, in 2022, researchers  officially completed one of the biggest   scientific undertakings in human history.  For the first time, we’d assembled an almost  complete list of every letter in human DNA. They’d been working on this for decades,  and it was like they’d finally unlocked   the instruction manual that made humans tick.

The possibilities for medicine seemed endless!  Except, there was a catch. No one was fluent in the   language this instruction manual was written in. Turns out, it’s one thing to know an organism’s   DNA sequence and a whole other thing to  figure out what those building blocks do.

And this isn’t exactly a  code you can crack on paper.  Because if you were to print out the roughly  3 billion letters stored in your DNA, they   would fill hundreds of thousands of pages. And cracking the code requires comparing   the DNA sequences of many individuals. We’re talking an absolute ocean of data.

Enter bioinformatics, again!  Scientists can use computers to compare  a bunch of DNA sequences until a pattern   emerges that lets us predict how certain  genes are working in an organism’s DNA. And these patterns can help us understand  and treat diseases, including cancer.  For example, there’s this gene called  BRCA1, which codes for a protein that   keeps cells growing normally, suppressing  tumors and potentially cancerous masses.  Which is a pretty big job! In fact, if even a single letter   in someone’s BRCA1 gene is off, it can lead to  an increased risk of breast or ovarian cancer.

Overall, variations in this  gene are linked to a chance of   developing one of these cancers by age seventy. But not every variation carries the same risk.  So, some scientists have turned to bioinformatics   to try to understand how dangerous  each of the gene’s mutations are. They compared more than 3,600 BRCA1  variants — trying to figure out what   the gene variants look like, what they  do to the tumor-suppressing protein,   and what overall effects  that might have in the body.

Scientists then compiled all of this annotated  data into a bioinformatics tool for others to use,   and the information it contains helps  doctors make more informed treatment   plans, which could mean better  outcomes for the patient overall. Bioinformatics can also help us  sort through which genes are being   expressed, or turned on, in an organism. You see, while some genes may be present,   only the ones that are expressed result in  protein production that impacts the organism.

To figure out which genes are being  expressed, we can look at all the RNA   molecules in an organism’s cells at any  given time, called the transcriptome.  This is an indicator of which genes are  being transcribed and used to make proteins. As with DNA, it would be impossible for a human  to analyze and compare organisms' transcriptomes   manually because they’re so complex. But with bioinformatics,   we can figure out which genes are being  expressed to make certain traits happen.  And if those traits are beneficial — like say,  disease resistance in crops — we can learn how   to use genetic modification techniques to  make other organisms similarly resistant.

Bioinformatics has super broad applications,  in nearly all areas of biology.  For example, it can help us gain a  better understanding of evolution.  By tracking how genes differ among closely-related  organisms — like different species of birds   or primates — we can gain insight into  how these organisms might have evolved. And, we can study random changes in  organisms’ DNA, called mutations.  Which not only helps us understand species  evolution better; it can also help us see   how viruses spread through a population. And that can be lifesaving information.

For instance, in 2016, the  mosquito-borne Zika virus   swept through the Brazilian state of Minas Gerais. And although Zika virus disease usually has mild   or nonexistent symptoms, it can lead to problems  with brain development in newborns or neurological   diseases, like Guillain-Barre syndrome, where  the body attacks its own nerves, in adults. So, an international team came together to  use bioinformatics to try to understand how   this virus got to Minas Gerais and how it  evolved over the course of the epidemic.

The researchers took samples from  patients with Zika virus disease.  And then, they did something pretty amazing:   Right on the spot, they were able  to analyze the DNA in those samples   using handheld nanopore technologies. Which is about as sci-fi as it sounds. These are essentially tiny computers that let  scientists sequence—or figure out the precise   makeup of— DNA and RNA, wherever they are.

There’s no need to ship anything to the   lab — you just feed the device a  few drops of sample and hit go.  Then, those sequences were uploaded to a  larger database, and compared with each   other to build a picture of how they were  related, a little like a virus family tree. Ultimately, this DNA analysis told the team  that the virus had been circulating in Minas   Gerais for at least sixteen months  before it was confirmed in the lab.  And by gaining a deeper understanding  of the outbreak’s journey and timeline,   researchers were better equipped  to slow the disease’s spread. [CHAPTER 8 - BIOINFORMATICS & PROGRAMMING] No matter what you’re studying, with bioinformatics, the computer  is doing a lot of the math.  So while it’s important to understand  what an algorithm is trying to do,   you don’t necessarily have to be an  expert in math or programming to use one. That said, despite being able  to sort through piles of data   faster than you can say “bioinformatics,”  algorithms aren’t all-knowing overlords.

You see, algorithms work precisely because  programmers give them limits and assumptions,   which are set values that  the system assumes are true. For instance, if we were to create one  for our pizza survey, we might give it   the assumption that anyone who didn't  like their pizza folded ate it flat.  This helps it make calculations  more quickly because now it knows   that any answer that doesn’t fit  the “folded” value must be “flat.” Limits and assumptions are really important,  and if you don’t know what they are,   your program might give you misleading  results, or might not work at all. Like, say I was sorting  through that pizza database,   and I asked a computer to give me  all the data about Deep Dish pizza.  If it was programmed to limit  results to just data about topping   preferences, it might throw me an error. “Sorry, Dr.

Sammy. Deep dish pizza doesn’t exist.”  But that’s not true, I’m just  looking in the wrong data set. So, just like a lab coat, a clipboard, and  a microscope – computers and algorithms are   tools in the biologist’s utility belt.

They can be used to analyze huge sets   of information that would take us humans  many lifetimes to work through on paper. Thanks to bioinformatics, the fields of  biology have advanced by leaps and bounds,   and continue to grow as engineers develop  better computers, and programmers build   better algorithms that allow biologists to ask  more and more complex and fascinating questions.  And speaking of complex questions, next  time we’re going to answer a weird one:   why aren't we made up of just one big cell? I’ll see ya then!

Deuces! This series was produced in  collaboration with HHMI BioInteractive.  If you’re an educator, visit  BioInteractive.org/CrashCourse for   classroom resources and professional development  related to the topics covered in this course. Thanks for watching this episode of Crash  Course Biology which was filmed at our studio   in Indianapolis, Indiana, and was made  with the help of all these nice people.  If you want to help keep Crash  Course free for everyone,   forever, you can join our community on Patreon.