Personalized medicine

A website where oncologists can upload their patients' HLA types.
They receive a list of peptides that may be used as a vaccine.

Similar projects worth following
Our ultimate goal is that oncologists become able to design custom peptide vaccines in minutes.
This web service illustrates this vision. When filled in with relevant information, it is able to design multiple peptides with different profiles (better fit for a given HLA, most immunogenicity, etc.).
This Proof-Of-Concept is obviously a work in progress that does not pretend to have any medical value.

If oncologists (or their health providers) were able to define vaccines in either the off label use, compassionate or expanded access legal frameworks, lot of people would be helped.

We are focusing here on peptide vaccines because while most of the clinical results using peptide vaccines have been disappointing, a peptide vaccine with appropriate adjuvants is easy to produce, low cost and probably one of the easiest strategies to induce tumor-responsive T cells in people who are still in good health.

The mammalian organisms have all many tumors as their cells, which are incredibly complex little machines, reproduce themselves often incorrectly so there is the need of quality mechanisms to remove those many aberrant cells. The immune system is one of those quality system, the check points in the cell cycle are another. Even in DNA replication there are quality controls!

Human leukocyte antigens are molecules on the surface of cells that allow their identification by the immune system. They are the marker of the "self" as opposed to foreign materials. These proteins are generally 7-11 amino acids long and come from proteins expressed by the cell. Under normal conditions, cytotoxic T lymphocytes, which normally patrol the body in the blood, and "read" the peptide presented by the complex.

T cells bind to foreign peptides and a series of events is launched culminating in cell death via apoptosis. In this way, the human body eliminates cells that are expressing foreign proteins.

To take advantage of this mechanism, vaccine immunotherapy administrates ad hoc proteins that mimics proteins expressed by  a specific cancer cell.

Those small proteins, called peptides, are synthesized by laboratories and then injected into the body as a vaccine to try to induce an immune response against the cancer cells.

Designing those peptides is not easy, they must be tailored to the patient. They must be effective, close enough to a protein produced by the cancerous cell to make the immune system to kill those cancer cells when it will encounter them.

Unfortunately those vaccines are useful only for a short time as the cancer cells evolve quickly. Another thing is that cancer cells can recruit the immune cells for their own purpose or make normal cells in the surrounding extracellular matrix to silence the immune system cells.

It might seems that cancer is a terrible, incredibly foreign thing, but actually it might simply be that it is merely the tissue's healing process that went rogue.

Another aspect that might ruin the effectiveness of the best vaccine, is the magnitude of side effects. Side effects should be expected as pathways are never specific, they are used in several purposes. Side effects might be so hard to endure that the patient might decide to leave the treatment, or might die from it.

We wrote a PoC of a peptide vaccine design web service. The code is available on Github at

And a demo can be played at:

  • Some preliminary code

    Jean Pierre Le Rouzic01/24/2018 at 13:17 0 comments

    This is a wrapper around NetMHCpan and a few other university services like VaxiJen.

    It tries to keep things simple, the user enters her gene, her mutation and HLA type and the service proposes peptides.

    I am not sure it is very wise in its design and intended purpose, so I have other ideas.

    Keep listening....

  • A service to find which peptide corresponds to which HLA type

    Jean Pierre Le Rouzic01/15/2018 at 16:56 2 comments

    When scientists want to design a new vaccine against some disease they can use the old pathway: Use some weakened form of a pathogen to stimulate the body's immune system to recognize the agent as a threat, destroy it, and keep a memory of that pathogen, so that it can recognize and destroy it with much more efficiency if it  encounters it in the future.

    It works because some peptides of the surface of the pathogen are recognized by the immune system. Those peptides are called epitopes.

    Nowadays, scientists prefer to design a vaccine by designing a peptide which is similar to a given epitope. The main roadblock is that different people have different immune system "sentinels". Some of those sentinels that patrol tissues are called MHC, they are proteins that have on one side receptors tailored to recognize a peptide decorating a pathogen, and on the other side they can activate T-cells (a subtype of blood's white cells).

    It exists many services that help scientists to design peptides for a given MHC type, for example:

    However it is quite complex to use. One problem is that one must seed NetMHC with another peptide in order to receive the prediction from NetMHC. the choice of this peptide is not trivial.

    A much simplified service, tailored for cancer vaccines and still relying on NetMHC 4 is available here:

    PS: Indeed it is not enough to have this peptide to make a vaccine, one must add an adjuvant that will trigger the immune system, as the peptide most probably will be completely innocuous. The choice of the adjuvant and its administration is itself quite complex.

    Another thing is that playing with the immune system is playing with fire. For example many recent drugs like check point inhibitors create auto-immune diseases like diabetes. Going from Charybdis to Scylla?

  • Is this serious?

    Jean Pierre Le Rouzic11/08/2017 at 08:15 0 comments

    There are many web services that tell which peptide fits the best to a HLA type.

    Indeed the first step in creating a personalized vaccine should be to choose the peptide that will trigger a response by the immune system.

    * One problem is that there is nothing as "a response by the immune system":
            Because there are several immune systems.

    * Another is that there is nothing as "triggering a response".
            The same peptide could trigger an immune response against cancer cells, but it could actually trigger a response that promotes cancer.

    * The immune system has mechanisms to block unwanted immune responses, the famous "check points" that stopped cancer immunology to progress during decades. True there are "check point inhibitor" drugs, but using them is playing with fire.

    * HLA typing is not even close to being "personalized". HLA typing as done by the industry, will search for less than 10 major HLA types.
            HLA types are personal, they are "genetically highly variable" and it is what makes each of us different from the next person. We evolved this way in order to make spreading of a contagious disease difficult.

    * Cancer is a tissue disease, before being a genetic disease. Something that inhibited cancer immunology for a long time was the fact that the immune system could not penetrate inside tumors (for complicated reasons).

    * Even if we can find the best peptide, there will be lots of side effects. Current web services give a molecular answer to the question, not a medical one. Actually they answer to the question: "Which peptide docks the best on a conventional HLA type". There is nothing about the patient and her side effects. Side effects could kill the patient quicker than her cancer.

    * Molecular docking is an insolvable problem when using classical computation. Maybe quantum computers will change that, but for now there is no quantum computers. So how work those web services? It is simple, they train a Machine Learning algorithm (the kind we use in our Heart Failure detector) and the algorithm can easily spit out some answer. Each academic that wrote such a service dismiss other proposals, but is unable to prove that her ML algorithm is better. ML is a dirty process that should be used for things that are not vital, or things that could be crossed checked many times. While this is what could be called "an educated guess", it is unethical to use it to create a vaccine.

    * If you want to have an immunology targeted treatment, you have to convince your doctor. It will be impossible to discuss with her if there is not a medical or scientific article that mention how this peptide could be useful for that kind of cancer. As far I know there is no peptide/HLA docking web service that do that.

View all 3 project logs

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates