Team:SYSU-Software/Human Practices
Overview
To make our project beneficial and responsible, this year we have conducted Human Practices in many different fields, such as biology, computer sciences, bioinformatics, diagnostics, and law. We have engaged in various groups, from iGEM team members, to college students and professors. Human Practices shaped our project in a profound and meaningful way.
Silver Human Practices
HP with bioinformatics professor
After putting forward the idea of our software and completing the demo, we visited Dr. Jian Ren majoring in bioinformatics. At the beginning of our project, we planned to use image recognition to let the machine directly recognize different shapes in the articles, But Dr. Ren told us that the current machine learning basically cannot do this, or the accuracy is not high enough. Therefore, we re-conceptualized and designed the OCR +YOLO4 position recognition algorithm as the final method of our pathway extraction, which was agreed with by Dr. Ren. Dr. Ren also provided his advice and guidance on copyright issues, webpage beautification and wiki writing in our competition.
Discussing with Prof Ren Jian
computer sciences
Professor Kun Zeng
Institute: Institute of Software Engineering and Application
E-mail: zengkun2@mail.sysu.edu.cn
Homepage: http://ss.sysu.edu.cn/~zk/
Professor Zeng's research interests include computer vision and machine learning, medical image analysis and financial big data analysis. Therefore, we believe that our communication with Professor Zeng will provide us with valuable opinions and feedback on our projects. In the form of interview, we learned about Professor Zeng's views on the future development of our project and the field of bioinformatics intersection.
First, we introduced the framework of our project. After asking some details, Professor Zeng Kun confirmed our idea and concluded:
1. Image searching is an interesting area to explore, and there is a lot that can be done in this area. The process of image recognition is very complicated, and it is not easy to be a good searcher because of many factors such as structure similarity and pixel similarity. In this respect, many software do not do well, they can only find completely consistent pictures, but it is difficult to find very similar pictures. Therefore, it is very breakthrough that you are willing to try this direction and have made certain achievements (accuracy rate near 70%).
2. However, I heard you say that the updated database has become smaller, with only a few hundred images. Such a database is too small for the software to be applied on a large scale.
Next, we asked Professor Zeng some questions about machine learning and biomedical research:
SYSU-Software
We see on your home page that you do research in medical imaging analysis. Can you tell us something about this field?
Zeng
I once cooperated with others on a tumor identification project. In general, when we get some medical images, a very large amount of data, it takes at least 24 hours for three or four people to manually examine each image to look for potential lesions. Using machine learning algorithms, potential lesions can be found in just a dozen minutes, and then the images of these lesions can be rated, ranked and given to humans for final identification. But it is impossible to give 100% of the work to AI, and there is a huge risk of failure and legal liability. The consequence may be that dozens of patients lost their lives because of misdiagnosis.
SYSU-Software
We have long dreamed of doing text retrieval of literature, identifying the underlying channel structure information, and helping to improve our accuracy in identifying the metabolism map information, but programmers tell us that it is difficult to identify text information from the document PDF. Is that true?
Zeng
Yes. Because some PDF files are encrypted and some are pure images, decrypting them and extracting the text is no less difficult than OCR directly using the image. For our existing projects, the proposed improvements are to increase the font size and add semantic analysis capabilities.
SYSU-Software
What do you think of the future of biology and informatics?
Zeng
In the past ten to twenty years, especially during the coVID-19 epidemic, we can see that informatics is the hottest and most promising subject today. In fact, in the past decade or two, with the exception of some areas of informatics and biology (such as genomics analysis), all disciplines have come to a complete standstill. Therefore, the 21st century will continue to be the century of informatics. You can't work in biology without working with computer science.
Consulting Prof. Zeng Kun
Questionnaire
To get a better idea of what iGEMer and synthetic biologists expect from computer software, we created a questionnaire to investigate in Chinese and English edition.
Our questionnaire surveyed a total of 49 Chinese students and professors in the field of synthetic biology, and 6 overseas synthetic biologists. Among them, there are 44 students and 11 graduates and research workers.
We first ask which of the challenges is prominent and urgent in synthetic biology?
The results are four questions that most researchers are more concerned with.
Secondly, we also put forward some questions about computer software aided experimental optimization. To see if the researchers believe and are willing to do experimental optimization in this unconventional way.
The result is below here
Then we asked a lot of questions about the functionality of the Maloadis project itself, starting with search, design, and so on.
For example, we ask interviewees to rank these situations from the one you want to improve the most to the least. And we receive the rank below.
After the investigation, we have a comprehensive understanding of the problems researchers are facing in the field of synthetic biology, and understand the problems that most researchers need to solve urgently. At the same time, we also know the interviewees' preference to use the software, like our maloadis!
Full result of our questionnaire check here.
Mini class
Theme: Synthetic Future: The romantic encounter between synthetic biology and bioinformatics
Keynote speaker Mo Ruoheng
In the spirit of promoting synthetic biology and iGEM, SYSU-Software held a mini class for freshmen to introduce them to this wonderful and vibrant area.
First，we throw a problem. What exactly is synthetic biology? Before answering this question, we envision a future in which synthetic biology will lead to a more convenient life, from the treatment of everyday ailments to the industrialized production of high-throughput products required.
And we emphasis that synthetic biology attempts to explore the origins of life, to study the mechanisms of life, and even to create life from nonliving matter.
The development of synthetic biology is also inseparable from the integration and promotion of bioinformatics. And bioinformatics is the study of biological problems using the methods of applied mathematics, informatics, statistics and computer science.
The combination of the two has produced a fantastic reaction and a intriguing journey from data mining to synthetic life with infinite possibilities.
Then we introduce the concepts of circuit, part and device in synthetic biology, discuss how to solve the actual problem, and how to standardize the synthesis process.
Finally, we also introduced our project maloadis to the freshment, hoping that they can understand and appreciate this wonderful melody.
Theme: Synthetic Future: The romantic encounter between synthetic biology and bioinformatics.
Integrated Human Practices
Biosafety
Biosafety has always been the first priority of scientific research, and the premise of all research should not violate ethics and social security. We paid a special visit to He Zhumei, a professor at the School of Life Sciences at Sun Yat-sen University. We consulted him about the bioethics and biosafety sections of our software.
First of all, Professor He introduced four basic principles of biosafety to us:
• The principle of respect:
Respect for whether or not violation of bioethics is harmful to human dignity and whether or not harm to other lives.
• The principle of beneficence:
Recognize that every study needs a benefit research team for the present and future benefit of the entire human society.
• The principle of no harmlessness:
Ensure that the work is not harmful to the environment and life.
• The principle of fairness:
Unfairness will lead to discrimination against human society and other social issues such as bioethical problems.
After the consultation we make a little change in our project.
First of all, our biology chassis is mostly E. coli, so the ethical issues are not prominent. The main things to think about:
1. Will other substances harmful to the environment be produced when the output of the target product is increased?
2. Will the process of synthesizing a pathway produce harmful substances that we have not been able to identify?
3. Influence of plasmids in the hosts.
For the teacher's suggestion, we added a section of the safety notice to the project to remind the user that everything should be carried out under biosafety conditions, and we especially emphasis the above contents.
Wet-lab Experiment Advice.
Professor Jianzhong Liu
Director of the Institute of Synthetic Biology, School of Life Sciences, Sun Yat-Sen University
Homepage：http://lifesciences.sysu.edu.cn/teachers/professor/248
First of all, we introduced the framework of our project. After asking some details, Professor Liu first confirmed the overall idea of our software and pointed out that search was a highlight of our project, but also raised some questions.
1) We need to build a large enough database of pathway diagrams to meet the needs of more synthetic biologists
2) We need to be clear about how search platforms help synthetic biologists solve problems
3) We hope that our search results can return the research interest and field of Daniel, so as to facilitate further research by synthetic biologists
Then, we asked Professor Liu some questions about the development of synthetic biology.
SYSU-Software
What are the challenges of synthetic biology?
Liu
I think first of all, how to translate scientific research into applications and create economic value is a major challenge for the development of synthetic biology. However, the links between synthetic biology and medicine, as well as issues in biosafety, are of vital importance.
SYSU-Software
What role will computers play in these challenges?
Liu
Big data platforms could be an important direction.
SYSU-Software
What are the common problems in synthetic biology experiments?
Liu
There will be many repeated experiments, and due to the complexity of biological metabolic networks, global regulation of cells will require multiple rounds of attempts, with uncertain results.
SYSU-Software
How can computers help solve these problems?
Liu
In the future, automated experimental platforms will greatly liberate our labor force and focus more on design innovation. Some algorithms can also help us to find the appropriate path adjustment points.
Bio company
Maloadis is an integrated synthetic biology software. In addition to being applied in schools and laboratories, we also hope that synthetic biology companies can pay attention to our software and evaluate the practicability of our software from the perspective of market, industrialization and commercialization. Therefore, after completing the software demo, we visited Guangzhou KingMed Diagnostics.
During COVID-19 epidemic, with personal protection, we visited KingMed on Oct 16th. Angie Shaw, the Securities Affairs Representative and Deputy Director of Board Office, received us.
First, Angie Shaw introduced to us the development history of KingMed, its main research fields at present, and its cooperation projects with hospitals. We discussed how the company built the database of diseases for Asians, and we also raised questions about security and privacy. Mrs. Xiao told us that the company's patient data were kept and reviewed by the ethics committee. The database was built with a collective sample and no information was disclosed to individual samples. We have also learned some relevant regulations on biosafety and ethics, which has guiding significance for our software.
We then discussed whether our software could work in this area. Angie Shaw told us that synthetic biology would have a great prospect in medical science. Synthetic biology studied the pathways in human cells through the modification of bacteria and viruses by artificially designed gene circuits, which realized the monitoring, diagnosis and treatment of certain diseases through disease-specific signals and specific targeting of abnormal cells. Our software had the function of pathway simulation and prediction, which was expected to predict the signaling pathway in human diseases through the pathway simulation in relatively simple prokaryotes.
Finally, we talked about machine learning and the big data used in our software. It is obvious that our future is the era of big data, which needs complex analysis to obtain meaningful information from numerous data.
Therefore, both machine learning and software development are necessary, which can reduce the artificial time and improve the efficiency. In the medicine field which KingMed focused on, it had developed rapidly from traditional medicine to precision medicine, which analyzed specific disease types of such as cancers and accurately find drug targets through genome, proteome and other omics technologies.
Through genome, proteome and other omics technologies, specific disease types of large sample population can be analyzed, identified and applied for biomarkers, so as to accurately find drug targets. Therefore, it is essential to combine biological big data, information science and the disease mechanism to achieve targeted, real-time monitoring and integration of biological big data.
precision and accuracy of medical is the essence of genome and proteome omics technology, particular for large samples to analyze biomarkers of disease type identification, application, thus accurate search for drug targets.
Intellectual property
We consulted Guang Dong Ying Zun Law Firm about intellectual property in our project.
SYSU-Software
Our database needs to store a large amount of literature illustrated pictures, and the images can generally in the open literature or purchase in the literature, which Sun Yat-sen University has mostly bought. Meanwhile, we will transform the image to our new databases after dealing with the program. When using our software, users can directly obtain the extracted information, the original legend, and the literature name corresponding to the above information. Is there any illegal use of literature in this process? And what we need to supplement and pay attention to in this process.
Law Firm
1. If the above is used for academic research and communication, it is ok, but the source should be noted.
2. If it is developed as a business and makes profits from it (i.e. operating activities or fees), the consent of the original literature author and written authorization shall be obtained. If the fees are to be paid, the fees shall be paid before use.
SYSU-Software
We integrated three databases on the network in another part of the project:
RegulonDB (http://regulondb.ccg.unam.mx/index.jsp)
RegulonDB's data were obtained via direct download. Is there any illegal use of the database in the above procedure? And what we need to supplement and pay attention to in this process?
Law Firm
The same as above 1, you must also obtain the data owner's consent and the written authorization permission to use. Paying the expense for usage if it's necessary. And the copyright owner of the database should be found in your work.
SYSU-Software
In the designer part of the project, we will provide functions for users to upload and share their design results, as well as browse and download other users' design results. These designs are based on the basic components provided in our designer. How should the user protocol on this function be written? And what we need to supplement and pay attention to in this process.
Law Firm
An agreement shall be signed between the user uploader and your platform, which shall stipulate:
1.the users to upload voluntarily allow other uploading browse and download to use. And the progress has no definitely relationship with your software (in fact like a Shared platform). The sharing platform does not assume any legal responsibility for the content uploaded by users and uploaders, but only acts as an intermediary.
2. The copyright of the design work uploaded by the user uploader is suggested to be owned by the platform, and the platform has the final right of interpretation.
The chat record of our consultation with Lawyer Yang
Development
After doing researches in Human Practices, we realize our project goal should help facilitate the development of synthetic biology. From our questionnaire, we learned that we can do it in two ways:
1.Promote standardization:
Users can search both genetic circuit images from published papers and standard genetic circuit images drawn with parts from iGEM Registry. In our design platform, we use standardized SBOL icons to help users design. In this way, more synthetic biologists will learn and understand the benefits of standardization.
2.Improve efficiency of design:
We added a project goal of design automation, so that users can get a design simply by setting their input of demands. Besides, they can also use our image search to search for matching genetic circuit design in a fast and convenient way.
For more information, check here.
Questionnaire result showing the importance of standardization
