Failure is not the end. It is a constant companion!

A short PhD student journey

If we look at people’s curriculum vitae (CV) /résumés and career paths, we might think: Wow…their life runs smoothly. However, what we don’t see, and what, of course, no one usually mentions, are the many mistakes and failures.

Let’s just dive right into one chapter of my life journey when I started a doctoral thesis (PhD) in neuroscience.

You must understand that the success of neuroscientists is historically and still nowadays evaluated by how successful they are in publishing their work in international journals. By now, I have two first-author publications and five co-author publications. Notably, the science community evaluates publications where you are a first author higher. It assumes that you have made a larger contribution to the work.

Besides publications, my CV/résumé shows continuous education without any gaps, finishing studies the regular time, many internships and lab rotations, conference talks, and a good amount of extracurricular work. It looks like it went all smooth, and there is sufficient output of invested work. I am not mentioning this to brag, but to take this as an example that the real story behind something that looks very smooth is completely different, and I am sure this is the case for many people.

In this story, I will describe how I failed many times and how things didn’t work out as planned. Something that you cannot find in people’s résumés. Therefore, you will get a look behind the curtain of the good-looking résumé/CV. The neuroscientific research accompanying this story will be explained in a way that everyone without any neuroscientific background can follow.

Starting my PhD, I have worked on a project developing and testing a newly developed electrode array to record from the surface of the cortex. This device features planar electrodes arranged to fit the size of tiny rodent brains embedded in a thin insulator. The main characteristic of this electrode array is its transparency, allowing for simultaneous optic experiments and electric recordings. It is possible to apply neuromodulation techniques like light-induced modulation of neural activity via optogenetics (see my previous post here) and, at the same time, record electric neural activity. I tested those electrodes by recording signals from the part of the brain that codes for auditory and visual stimuli. Since the array is mainly transparent, I applied laser light through the electrode array while also recording the neuronal signals. This allowed precise modulation of the neural activity.

So far, so good. Time has passed. We have done a lot of different experiments to test those electrodes and improve their features, for example, making them more flexible to reduce neural tissue damage. And, of course, the plan was to use those electrodes further.

Our plan was as follows:

In our lab, we focus on a neurotransmitter called dopamine which is important for many processes, including plasticity. Neural plasticity is required for learning and memory. Neural plasticity simply describes the capability of modifying neural activity and that a neuronal network can change depending on the experienced environment. In simple words, dopamine takes part in having a flexible brain that adapts to a changing environment around us.

We recorded signals from the auditory cortex when animals learned an auditory behavioral task. Later we wanted to modulate those signals by activating specific neural projections that release dopamine. One can better understand the role of dopamine in learning and memory by releasing it at very precise specific time points during a behavioral task.

One requirement to observe changes in cortical activity is that the recorded signal itself is stable over time. While the recorded signal amplitude of our developed electrodes is generally fine for many months, there is one problem.

I never managed to get a stable resolution of the auditory cortex tonotopy.

What is the tonotopy? Each auditory frequency has an area of representation in the auditory cortex. A 2 kHz tone activates mainly a different area in the cortex than a 10 kHz tone. The stability of tonotopy is crucial because we want to see how these tone representations change over time during specific dopaminergic modulation.

BUT we failed! This resolution always vanished over time.

Figure 2: An example of something that has failed. Here you see the activity of the auditory cortex measured with electrodes while the animals listen to two different tones. While the two tones clearly evoke activity in different areas initially, this clear separation vanishes over time until one cannot say anymore that one tone activates an area of the cortex that is different from another tone.

We simply could not get a stable signal of the tonotopy for the necessary amount of time (months). This is, by the way, not uncommon with these types of surface electrodes. Later I found out that no one had managed to get these stable tonotopic recordings for such a long time. But science is always about trying out what has not been possible before and advancing, so we gave it a try.

We tried to resolve this problem with many troubleshooting attempts that included different surgical procedures to minimize the neuronal damage introduced by the implantation surgery and changes to the recording setup. But it didn’t work. As if this is not enough, other problems appeared. For instance, the supply of these custom-made electrodes was not stable and sufficient then. We couldn’t just buy them for a company but collaborated with engineers to produce them. They had difficulties with the fabrication process, and other political reasons at the university slowed the process.

The behavior setup, which was also needed to advance this study, required a lot of very advanced programming. However, the collaborations with other scientists to work on these programming steps were falling apart. Additionally, the project planning and management from a supervision side were also not the best.

A side note:

You must know as a PhD student; you are often on your own. People tell you to get something to work, and you must figure it out yourself. So often, it feels overwhelming.

Another side note is crucial to mention. By no means does this text imply blaming anyone for the project’s failure. It only shows in some snippets from one perspective how things were going or felt at that time. This includes own mistakes and external not controllable factors.

Coming back to the main point.

What happened next?

We successfully published this electrode array for cortical surface recordings, which is still a valuable advancement in that field. The electrodes we developed to record neural activity still have many advantages over previous ones and are generally working fine and can be used for many neuroscience applications.

However, the amount of time and effort invested in our planned project (using those electrodes in the auditory cortex and doing dopamine-related experiments) that ultimately failed is very frustrating. It meant my planned PhD project had failed. It leaves a feeling of not knowing what to do now. Everyone sees the success of the first part of the project (the electrode array that works) but not the failures of all the successive projects we had attempted. I am talking about years of work from which only the initial part of a project was successful, but all the other parts ended up in nothing fruitful.

I want to mention that I describe a part of my story to show that failure is normal. And I know everyone has a similar story about the failure of their plans and projects. That can be both privately and in a professional context. This story here is nothing more special than your own.

Maybe it is important to share some of them so that people starting a PhD or any other undertaking don’t believe that their failed projects are the worst thing ever. But that they see it is normal for everyone to experience major failures.

Let’s talk more often about the things that didn’t work out. Because I have the feeling that if people go to scientific conferences, they see all these fantastic talks about what other scientists have done, and it all looks very smooth and easy. The same goes for reading scientific papers, which nowadays often promise to have found the solution to a problem and did so without any flaw.

But you must understand that these talks and papers are polished versions. Why? Because it seems easier to tell a story that is purely successful than a story that includes mistakes and setbacks.

There is always a story behind the story. Talk to those people, and if they are honest, they will show you their mistakes or setbacks. Such things are not so easily visible at first glance.

There is more.

After the first project could not be continued, we started something new. And that time, more than half of my official term as a PhD student was already over (meaning some years had passed).

We still wanted to work with dopamine, but this time focused on the dopaminergic neurons in the ventral tegmental area (VTA). The area of the brain where most dopamine neurons have their origin. To study this, I built and implemented a new behavioral task in our department, a reward-based task for mice. While I made many mistakes in building this setup, it worked in the end. So far, so good. Simply put, we let the mice perform a learning task and wanted to see what dopamine neurons were doing. But this area of the brain we are looking at also has other types of neurons. This means we needed to identify dopamine neurons, among other types of neurons.

One can use a technique called optotagging to send a short light pulse into that area, which as a result, activates neurons that express a light-sensitive protein. We only made dopamine neurons light-sensitive with a genetic modulation approach. Other neurons didn’t answer to that light stimulus because they don’t have the protein that is light sensitive. This allows us to tell which of the recorded neurons is presumably dopaminergic and which is not.

Long story short. It didn’t work out for months.

We got help from a well-known lab at Harvard, and for them, it also didn’t work very well (the yield of finding such dopamine cells is very low). Again, months of troubleshooting and finding a possible solution. In the end, I can say that we found some major factors to improve this technique. This includes using a different mouse line that has a more stable expression of that light-sensitive protein expressed in dopamine neurons. However, by that time, I had already invested a lot of years into my PhD and project funding came to an end. This meant I had to finish up my work.

Despite that second major failure in not being able to finish the second PhD project in time, we published a newly developed device that we used to record from those dopaminergic neurons. The important message here is that even if things don’t work out as planned and there is a lot of frustration, it is possible to adapt and make something out of it.

We published work that we developed for this project, an open-source recording device that can be 3D printed in other labs. It allows other scientists to use this device to progress faster and easier in their respective projects. However, everyone again sees the success of the second publication and not the failure of a much larger project behind it.

Ultimately, I would like to say that it can be tough not to see these failures as failures of oneself. First, every human makes mistakes, and it’s fine. The important point is to not repeat them many times. Secondly, there are also many external factors that you cannot control that will lead to failures. And the only thing you can do about it is to learn and be better prepared for the next time. Failure leads to something successful. It can be the learning process or progress and valuable knowledge for others. Because sharing what didn’t work is also a form of progress.

Keep in mind: no one has a résumé of all the things that went wrong, the prices they didn’t win, the applications for jobs they didn’t get accepted, the publications that got rejected.

The point of sharing mistakes and problems is not to tell everyone how hard and complicated everything is and not to complain about everything. But talking about failure and setbacks helps normalize them so that no one feels like a failure.

Last but not least, there is something to add

Maybe you feel or think the following now:

It’s easy to say that talking about failure is helpful, but what if others see me as a failure and, for instance, do not hire me. I agree with you that showing a weak side can be problematic in some circumstances. Angela Merkel, the chancellor of Germany for 16 years, only in her last year, when she announced already that she was going to resign, had a more emotional speech and also included talking about mistakes in some decisions. It included a story of personal struggle during her upbringing in a time when some parts of Germany still had a dictatorship. As a leader of a country, people usually do not show vulnerability because it can be used by an enemy or competitor against them. It’s a job of power. It’s a job where it’s expected to always make the right decisions, and a slight deviation is immediately attacked.

But we are not talking about such an extreme level. Before a change can happen at such a level, we must start with everyday life.

Talking about mistakes doesn’t make someone weak, but it is honest. If we normalize talking about failures and problems in our lives and someone thinks we are too weak for a job. Then let them believe that and walk away. In that case, you don’t want to work with them anyway. I think many of us live in a world where we have the privilege to choose (to some extent) with whom to work and with whom not to work. People should be able to see that you learned out of those failures or setbacks and progressed from that.

Accepting a person as a human with both positive and negative aspects is the key. No one is a machine and can just do everything right. And no one should expect another person the be perfect.