Team:Tongji Software/Implementation
-
How to use Phage-MAP
-
Who are the end users
-
How Implement in real word
-
What other challenge
IMPLEMENTATION
★How to use Phage-MAP
Phage-MAP designs a platform to visualize the relationship between bacteria and phages and recommend candidate phages for phage therapy. You can choose to input a name of bacterium/phage, PhageMAP will then automatically performs database retrieval, multi-model prediction, interactive web page display and other steps, to get phages/bacteria that are highly possible to interact with your input.
These functions are integrated into one website and are divided into three modules: Bacteriophage Bay, Phage Finder and Interactive MAP. Clicking here Phage-MAP webpage, you can access the home page of the project. On this page, users can get help from the guidance videos and guidance manuals of our project. On the same page, you can also click to enter our three modules for the subsequent operation.
Phage-MAP website guidance video
★Who are the end users
1. For synthetic biologist
Phage engineering: We provide phage-bacterial interaction networks, which can help synthetic biologists artificially engineer phages to obtain broad-spectrum artificial phages, or synthesize new phages to enhance the advantages of existing phages, make up for their shortcomings, and further promote the engineering and application of phages.
2. For environmentalist
Advancing new treatment for superbugs could help reduce the amount of antibiotics in the natural environment, thereby reducing man-made damage to the ecosystem.
3. For patients infected with antibiotics
Phage therapy is effective and has fewer side effects than antibiotics against superbugs and healthy patients. With the help of Phage-MAP, doctors can quickly find suitable and efficient phage reagents to treat patients
To be continued ...
★How do we implement our project in the real world
After seeking for all kinds of Score, we build a model for predictions of phage-bacterial interaction. Our goal is to improve the current lack of data on phage-bacterial interactions and provide a reference for phage therapy, so that people will not be overwhelmed when applying phages. In the real world, researchers or medical professionals query the target bacteria in Phage-MAP, then Phage-MAP recommends the phages that have the strongest interaction with it. To ensure the effectiveness of phage therapy, the resulting phages are clustered according to their genome sequence similarity. Researchers can take a certain number of phages from each cluster, and then combine them together for cocktail therapy. We provide a range for phage selection to avoid users performing too many ineffective experiments to screen phages.
Despite a series of validation tests, our project is still based on computational simulations. It has difference with real experiments anyway, so all results from PhageMAP are for reference only.
★What other challenge do we need to consider
Though we have contributed our own solutions to the problem of phage selection and settled some obstacles, we still need to consider the following issues:
Limitations of genome sequence-based approaches
Our project initially started with CRISPR, hoping to determine bacteria-phage interactions by spacer alignment. But we soon realized that the CRISPR system only existed in a small percentage of bacteria, and known spacers that can be aligned to phage genome are far fewer. After discussion and communication, we decided to add alignment-free methods to complete our model, ensure that our phage-bacteria interaction covers as much phage-bacteria interaction information as possible. Although the model ended up making decent predictions, there are actually far more phage-bacteria interactions in nature than we thought. It is possible in principle to judge interactions purely from genome sequences, but it is only one aspect, plenty of important information may be unconsidered, especially when almost all of our methods are based on unannotated genomes (pure sequences). However, due to the extremely imperfect data in the microbial field, it is difficult for us to obtain all the detailed information about tens of thousands of bacteria and phages simultaneously, such as protein interactions, metabolic pathway interactions, and so on. If we add these, our prediction of phage-bacteria interactions will be more accurate.
Difficulties in result validation
At the same time, the lack of phage-bacteria interaction data also brings difficulties to data verification. At present, almost all known and confirmed bacteriophage - bacteria interactions came from experiments. However, the large number of bacteriophages and bacteria in nature makes it unrealistic to verify the interactions one by one via experimental methods. What's more, knowledge about phage and bacterial species are too limited, and many microbial sequences derived from metagenomic analysis have yet to be identified as specific species, let alone isolating and purifying from the environment for phage therapy.
Bacterial resistance to phage therapy
The generation and resolution of bacterial resistance is an eternal topic, it appears in every treatment options for bacteria, including phage therapy. Due to the high mutation rate of bacteria themselves, bacteria can produce mutant strains that can adapt and survive in almost any environment as long as the population is large enough. The single-receptor-based relationship between phage and bacteria seems to make bacteria easier to resistant than broad-spectrum weapons such as antibiotics. So theoretically, fixed-mode phage therapy might fail faster than antibiotics. But fortunately, Phage cocktails avoid this problem. In addition, advances in synthetic biology have made it possible to edit synthetic phages artificially. Alternate a phage’s construction by genetic engineering is much easier than researching brand-new drugs de novo for drug-resistant bacteria, this is also a promising solution to bacterial resistance to phages.
Tongji University, Shanghai, China
