Brainoid

When she went back that night, Qi Min sent the project plan she had written to apply for funds to Mr. Zheng Wenyu's mailbox.

The next day, when Qi Min was using the supercomputer at her school's computing center to continue running a larger test set with her existing miniature version of the "brainoid", she suddenly received a message from Zhou Xiaoshan, a researcher on the "Nimbus" supercomputer project of the Chinese Academy of Sciences. s mail.

The email told Qi Min that she would be able to go to the "Yuyun" computing center in Fangshan District tomorrow to discuss usage matters.

At this moment, Qi Min really wanted to kiss Yan Zhaolan - it turns out that old friends are more reliable!

Now there is no need to run this test set. Qi Min stopped the code for testing the learning ability of the brain-like body on Jupyter, and quickly started to go through the code for generating the brain-like body again. She renamed the variables that had been randomly named in a hurry according to the abbreviation of their purpose, and then wrote comments for each piece of code.

Qi Min used her "beautified" code to re-run an ultra-miniature version of the brain-like body with only 16×16×16 virtual neurons. She checked and confirmed that it was correct before saving the code.

If researcher Zhou Xiaoshan agrees to use the computing resources of "Nimbus", then she can try to generate full-scale brain-like bodies of 1:512, 1:64, or even 1:8 or 1:1. Once she gets a 1:1 brainoid, she can finish her article by just scraping together experimental data.

——

Qi Min got up early the next day and took the subway directly to the "Nimbus" computing center without going to school.

"Nimbus" is composed of a large computing core group. For heat dissipation considerations, it occupies a separate computer room. The upper floor is equipped with a complex interlaced liquid cooling pipeline system. It is obviously impossible for the staff of the computing center to let an outsider like Qi Min enter the computer room. Even the researchers working in the computing center rarely have close contact with real "rain clouds".

The place Qi Min was going to was researcher Zhou Xiaoshan's laboratory, and they made an appointment to meet there.

However, when Qi Min arrived at Zhou Xiaoshan's laboratory and asked the graduate students working there, he was told that their teacher was in a meeting. The graduate students in the laboratory did not ask Qi Min what he was doing, but Qi Min just went in and waited for a while.

Qi Min entered their laboratory and found that although it was also a bioinformatics laboratory, it was very different from what she had imagined.

The biggest difference is that there is no division between experimental area and living area in the laboratory. Because bioinformatics research requires very few "wet experiments" that operate on organisms or biochemical systems, there is no contamination from harmful reagents and microorganisms, so there is no need to separate the laboratory bench from the desk. Also corresponding to the characteristics of bioinformatics, everyone in the laboratory has at least two screens and unusually large hosts standing on their desks. The air conditioner in the room is turned on very high. Qi Min has no one on one of them. But the screen that was still always on showed that they were doing two-dimensional Fourier operations on a large number of electron microscope photos.

Qi Min found an empty seat and sat down, and saw a young man wearing Sennheiser headphones next to him who was debugging code.

She couldn't help but glance at it curiously and found that what he was writing was an RNA structure prediction tool. In fact, RNA structure prediction is already a relatively well-researched field. The existing algorithms have their own merits. However, for a specific type of RNA, it may be necessary to design a suitable special algorithm based on its known information.

This code looks like it was written by this young man himself, because Qi Min has been watching him modify it for almost ten minutes, and the output prediction array is still a bunch of negative numbers.

It happened that Qi Min had originally written the code for RNA structure prediction. Now he followed his modifications and read it, and he quickly found the problem.

"Hmm...this traversal method doesn't seem right." Qi Min was embarrassed to tell him directly, and just sat beside him and muttered.

At first, the young man pretended not to hear her while wearing headphones and ignored her at all.

Qi Min saw that it was really difficult for him to adjust like this, so he couldn't help but said: "For the nearest neighbor algorithm (i, n-i), we should start from i greater than or equal to 2. When i is 0 and 1, we must use the principle of the most hydrogen bonds..."

"It's the 'minimum energy' principle." The young man who adjusted the code took off his headphones and turned to her and said, "But the experimental methods of the initial measurement values ​​are different, and the two algorithms cannot be used universally."

"Who said that? The assignment of the first pair of hydrogen bonds has no effect at all. Just try it first." Qi Min said to him.

The young man doubtfully added a section of IF and copied the relevant code from another algorithm that simply used the principle of minimum energy.

After he ran it again, the output matrix was indeed not all negative numbers.

But obviously the result is not right either.

There are still too many negative numbers, resulting in the maximum value being less than 4, and it is obvious that the 30-base linear RNA used for testing cannot form only one base pair.

"What's going on?" The young man asked her directly this time.

Qi Min came to the screen and took a closer look, and then said while typing on the keyboard: "You didn't add FLAG to the nested FOR statement comparing the maximum value, just add it..."

The code was run again, and this time the output matrix finally looked normal.

The young man breathed a sigh of relief and was about to stand up from the chair when Qi Min said: "Hey, this is still wrong. Look at the known structure of this RNA, there are three LOOPs, but your algorithm calculated that there are four A... Oh, I know, I need to add an IF to prevent invalid pairing. Pairings of less than two consecutive bases should not exist."

After she added this IF statement, the final RNA secondary structure she drew was finally the same as the known structure.

"Uh... you're really..." The young man looked at the image on the screen and smiled unconsciously.

Then he glanced at Qi Min and asked a little strangely: "Are you new here? I remember that we will not recruit doctoral students here this year - are you a technician?"

"No, I just came to see Teacher Zhou Xiaoshan for something." Qi Min said with a smile.

"What's the matter?" the young man asked, and then seemed to realize that this question was a bit of an invasion of other people's privacy, so he added, "Maybe I can help you too."

"Not sure." Qi Min moved away from his screen, "I came to see Teacher Zhou Xiaoshan so that I could use 'Nimbus'."

"Oh, that's easy to talk about." The young man immediately agreed, "If my boss doesn't agree, you can still use my account."

"This is still not good..." Qi Min smiled awkwardly. Running a brainoid takes up a lot of computing resources. If you secretly use the account of a doctoral student, it will definitely arouse the suspicion of the administrator.

"What's the point?" the young man said nonchalantly, "'Nimbus' has such a strong computing power that ordinary projects are not a problem at all. The boss just wanted to start some big projects to test its actual computing power. We were recruited as biology people to pick up jobs here and there. To be honest, I didn’t learn much about Python before I came here.”

I can tell. Qi Min commented silently in his heart, and then asked jokingly: "Then if Teacher Zhou Xiaoshan rejects me, will I really come to borrow your number?"

"Okay, of course." The young man said immediately, "Come and add me on WeChat. When the time comes when this RNA algorithm has a bug, I will have to find you again."

Biological science and ordinary wet experiments are simply not the same thing... Those who can master interdisciplinary subjects are the big guys_(:з」∠)_