Nowadays in academic biomedical research, young people spend long years- as many as 7 to 8 years- in postdoc training, working day and night to keep getting data for publication. The only reason they work so hard with so little pay is to get a PI position someday. They are the real engines that drive the academic research.
We know the new PI positions in biomedical sciences become more and more competitive. Lacking carrot for horses, more and more graduate students give up career path of academic research. They will spend less than three years and even only one year in postdoc training, and then jump off to industry, regulatory affairs, or consulting business, etc.
At some point, the wave will become tsunami, and all the biomedical lab will run to the cliff of research personnel sooner or later. Eventually, that will lead to the collapse of the current academic biomedical research.
This new reality is coming, and academic researchers need to prepare and adapt to it. We must embrace cutting-edge technologies to make the research more efficient, more practical in other fields, and prepare trainees for different career paths. We need to start today, otherwise the whole enterprise will see doomsday approaching.
So how could technologies enhance the efficiency of biomedical research? Currently, we are witnessing “high-dimensional data revolution” in biomedical research: technologies and computational power drive exponential growth of data collected at every biological level, from genes and single cells to populations and eco-systems. Data analysis can generate hypothesis rather than letting hypothesis guide experimental design. In fact, such new paradigm has prevailed in sports. Now we see every ballpark or stadium is equipped with whole arrays of camera to catch every move, position, and physical condition of the players in every game. The video is converted to quantitative, categorized data. They are rich in variables, measurements, and connections, so-called “high-dimensional data”. The data scientists run the dimension reduction to identify the critical variable, i.e. "drivers", in the system and make suggestion on the parameters to control the drivers. The results are sent to the coach's team to make strategies.
Biomedical technologies that collect high-dimensional data, such as NGS or spatial transcriptomics, are like the movement-catching cameras in the ballpark. Computational biologists, like the data scientists in the sports organization, analyze the data. The results inform experimentalists to design new studies accordingly. Therefore, a modern PI’s responsibility is to form a team like sports business: collaborating with technologists for data capturing, work with data scientists for computational modeling, and leading experimentalists like the coach in sport team to run the game. This is how the "post-modern" biomedical research should look like. Trainees in such a team can develop multidisciplinary skills and perform research work efficiently. The “sports team model” would be the escape ladder under this personnel cliff.