Laboratory of Quantitative Biology (Esposito's group)

Large Language Models: How Our Lab embraces AI

In our lab, we’re not just observing the AI revolution—we’re living it. We’re actively integrating AI, especially Large Language Models (LLMs), to fundamentally change how we do science. This isn’t about replacing your expertise; it’s about giving you superpowers to be more efficient, insightful, and impactful in your research, all while upholding the highest standards of research integrity.

Our AI Compass: Agency & Ownership

Think of AI as a brilliant, tireless assistant. But remember two golden rules that are paramount to legitimate and responsible use:

You’re in Charge (Maintain Agency): You are the scientist, the thinker, the decision-maker. AI amplifies your work; it doesn’t do it for you. You must always retain ultimate control and intellectual responsibility for your content and conclusions.
You Own It (Assume Ownership): Any error the AI makes is ultimately your responsibility. You must always critically double-check, verify, and ensure every single piece of AI-generated information is accurate, unbiased, and sound. Your reputation, and the integrity of our lab’s work, depends entirely on your rigorous supervision of all AI outputs.

Your AI Toolkit

The AI world moves incredibly fast, but our core tools right now are ChatGPT, Gemini, and NotebookLM. NotebookLM is fantastic because it sticks strictly to your uploaded documents, making its answers incredibly precise and reliable for source-grounded tasks. We’re also keeping a close eye on Microsoft Copilot—it’s set to become a game-changer, integrating AI directly into the Microsoft apps we use every day.

AI in Action: Making Research Easier (Responsibly)

Here’s how we’re putting AI to work across our lab, always with an eye on data quality and ethical use:

Smarter Writing: Beyond fixing typos, LLMs are like having a personal editor. Draft your text, then ask for specific feedback on clarity, coherence, or even if your manuscript rebuttal hits the mark. This helps you refine your communication without compromising your original thought.
Rapid Literature Reviews: Say goodbye to endless scrolling! LLMs can quickly identify, summarize, and analyze stacks of academic papers, helping you get up to speed on new fields or dig deeper into existing ones in record time. Crucially, always cross-reference with traditional databases like PubMed or Google Scholar to avoid biases, and meticulously verify all sources!
Boosting Your Code: For coding, LLMs are becoming indispensable. They can help generate code snippets, debug tricky scripts, and even act as instant manuals. This accelerates development, but always review and test generated code for accuracy and efficiency.
Accelerating Quantitative Work: We’ve seen LLMs condense weeks of complex mathematical calculations into hours. For physics, they’re like a super-smart search engine for equations or specific components for engineering tools.
Data Analysis (with a strict caveat): While we use specialized deep learning for image analysis, we do not use, and strongly discourage the use of, LLMs for raw data processing, statistical computations, or drawing direct scientific conclusions from datasets. LLMs can “hallucinate” with numerical data. Use them to understand methods or debug code, but perform actual analyses with dedicated statistical software.

The New Super Skill: Prompt Engineering

This is key to effective AI use. Prompt engineering is essentially how you “talk” to the AI to get exactly what you need. It’s about being clear, specific, and even telling the AI what “persona” to adopt (e.g., “Act like an expert editor”). Master this, and you’ll get far more accurate and professional results, drastically reducing AI “hallucinations” or overly flattering responses.

Remember: Humans First!

While AI is amazing, it’s a tool, not a replacement for people. Your interactions with colleagues, mentors, and peers are vital. Real human collaboration is irreplaceable for critical thinking, nuanced feedback, and true scientific breakthroughs. Scientific integrity is a human endeavour, supported by AI, never superseded by it.

Surprise! This blog post was generated with Gemini, using the original webpage I wrote to inform my colleagues about the use of AI in the lab.

Do you need a host for a MSCA fellowship (2025)?

Having re-established my laboratory at Brunel University of London, I am now open to supporting junior researchers interested in applying for independent funding through the EU MSCA scheme. I offer a uniquely interdisciplinary research environment, access to cutting-edge expertise and training, and a broad network of UK and international collaborators.

Brunel University is well connected to the European Union via London’s extensive public transport, the Eurostar, and nearby Heathrow Airport (just ~10 minutes by car). Our campus boasts state-of-the-art facilities in cell biology and engineering, further enhanced by my own research infrastructure.

If your background lies in biophysics, engineering, cell or molecular biology, computational sciences, or an affiliated discipline—and you see a strong overlap with my areas of research—I encourage you to get in touch. You can explore this website or review my publication record to assess the alignment. Please also consult the EU MSCA website to check your eligibility.

The deadline falls in early September, so I recommend reaching out in June if you’re interested.

The transition post…

A life between pandemic and departmental closure.

I resume writing after a while and just for a ‘short’ update. The last two-three years have been quite tough. Yes of course there was the pandemic. As tough as it was for most of us, I was lucky enough to have a relatively stable job and not feel the financial pressure of the pandemic in 2020.

Here is the point, ‘relatevly stable’ and 2020. Because of funding cycles, in academia we experience crises often with a delay. This is my experience, my story, my views, but there are many learnt lessons that could be useful for junior academics, post-doctoral scientists and students. I will thus share these with you but I needed time to move on, strip these lessons of emotional components and personal biases, simply to provide those couple of you who read this blog with some genuine advice.

Hence, this is a transition post, the only one that I will allow myself to share with more emotions, not just the passion with which I usually speak or write. I avoided sharing negative thoughts, and soon – very soon – I will share new positive experiences at my new job. But let me, give me, just this post…

While the pandemic was developing, our institute’s quinquennial review took place. While I was submitting documents under the philosophy of ‘business as usual‘, I was tracking students rushing home, who in the UK, who abroad. While I was revising documents apparently one page too long, I was seeing tragic scenes in my dear Italy, trying to keep in touch with friends and family who lived in the Asian and European epicentres of the pandemic. I was hearing about the first deaths in Cambridge, a nearby institute, someone in the neighbourhood.

While I was responding to referees, I was updating safety procedures and devices, trying to plan logistics for managing, procuring, trying to contribute to the community response, making sure my colleagues were safe and looking after my family.

Then the car rack that was the closure (no, not the lockdown, the actual closure, complete defunding) of our department with no certainty, little information, and deep anxiety in the second half of 2020. Eventually, we got a one-year extension, a rather standard support measure in a very non-standard period of historic proportions.

While the closure of a research institute, like the closure of a company, is absolutely legitimate, the measures that organizations deploy to mitigate the impact of their choices on their employees, speak a lot about the working culture of institutions. While the closure of a research institute is absolutely legitimate, it is much more difficult to think that that the closure of a publicly funded research institute during a global pandemic and without special mechanisms in place to mitigate the effects on employees could be done… honorably*.

A moment of crisis permits us to stress-test systems and to really distinguish between the fluff, the hypocrisy, the inefficiency, and tangible support for people. Of course, there is another side to the story. A moment of crisis permits us also to see how many good people there are, from your neighbour to your colleagues. Many helped, even just with a listening ear; a very few who should have helped, in all practical terms did not. Some that I would never have expected to help, even people that at the time I did not know well if at all, did help within the limit of what was possible.

Please avoid guessing who was bad and who was good because your guess will probably be wrong. People and organizations involved have reasons, have mostly good intentions. Unfortunatly, now and then, we slam against walls built of institutional bad habits, fatalism, and disconnect.

Although in this post I let myself to reveal a bit of the emotional rollercoaster that the last few years were, eventually my intent – my ambition – is to nudge a bit all of us to be better, our institutions to improve, and our working culture to progress. Think about an aircrash investigation** where assigning blame has modest consequences compared to the industry-wide safety benefits to avoid similar accidents to happen again. Eventually for the benefit of everyone.

Sadly, my repeated offer to analyse what has happened, not in public but as an internal procedure to improve in the future, has been so far dismissed.

My search for a new host organization was hindered by a job market affected by the pandemic, some risk-aversion in recruiting a physicist that does biology, or a scientist that does not have a formal track record in teaching undergraduate courses. And of course, by the psychological impact of living through this chaos which took its toll on the mental wellbeing of me and colleagues at any level. All of this is compounded by the shortcoming that I certainly have, like anyone else, of character and curriculum.

Despite the huge setbacks that my colleagues and I experienced, the delays in research productivity and the money virtually wasted in the process, I anchor myself to the understanding that most people in society and academia are good people, amazing and well-intentioned. I am indebted with those many colleagues in companies and academia who supported me. I will be forever grateful to the many kind-souls out there, that irrespective of their roles, position or means, did what they could to help.

I have one regret. I could not shield my team from all of that, as I did not have the power to. The only thing I could do is to to my best, including all my students and researchers, sharing information in real-time, discussing, listening, advising and supporting them within the limited resources I could tap into.

And now? Time to move on. Time to move to Brunel University London, where I found amazing people – soon to be new colleagues with whom we will do amazing things. But I will dedicate to Brunel ample space in the near future, focusing in what I predict to be an extremely positive experience in the making.

Time to move on. Well, I did move on in many respects, and left behind the worst emotions, the worst memories. I will however use these incredible experiences*** I passed through, to dig deeper into the mechanisms of academic life, as usual, trying to have a positive impact on younger scientists and students who might accidentally land on these boring, sometimes nerdy, pages.

I hope I will soon have time to write about the difference between power, leadership and management, the role of human resources in academia, the institutional bad habits that we should eliminate to ensure the good that is in most of us percolates up to our organization (yes, we need an antigravity machine of sorts), how careers between a research institute and University should be structured, or how openly speaking with colleagues in industry about career can be extremely positive.

I’ll have much less time with my new job starting, but I will have the right mental attitude to discuss those ideas without the shadows of the last few years.

Transition-post completed, the future is waiting.

*) I have been inspired by another event, another person, in the use of this sentence. And although only very few people will know, I still feel obliged to disclose this.
**) Not my philosophy, I hear this often in the Mentour Pilot Youtube channel. Inspiration can come from everything!
***) To be clear, other people in the UK and around the world experience much more tragic issues. I do not want to offend anyone sounding overdrammatic. Still, I can’t do anything else then telling my stories, lucky in a broader sense, hoping to be of use within my trade.

1D heat flow in a metal (mini-lecture)

I have recently presented a mini-lecture about thermal physics. It was part of an interview and rather off-topic relatively to my research. I had fun diving back into ‘pure’ physics and I thought to share the material I presented.

Lecture slides (pdf)

I also built and brought with me in the auditorium this 1m copper rod to visualize the ‘heat wave’ moving from the hot to the cold side of the rod.

Hyperlapse movie clip showing a copper rod heated at the right end. Heat flow is shown with a change of temperature visualized by a thermochromic paint, from green to red tones.

I was also able to shoot a clip with a thermal camera mounted on my phone. I did this really at the last moment in a hotel room (yes, I had to clarify I was not going to do anything strange in the room when I checked in with equipment 🙂

Hyperlapse movie clip showing a copper rod heated at the right end. Heat flow is shown with a change of temperature visualized with a thermal camera.

Finally, I prepared a handout. Again, this was rushed because I did not have much time and it was just a mock lecture, not a real one, but while I was making notes for myself while studying a topic I had studied a quarter of a century earlier, I thought to write them now.

Handout (pdf)

A virtual tour of my labs at the MRC Cancer Unit

All photos are 3D pictures, it was fun to try out the technology.

Regrettably, the MRC defunded the MRC Cancer Unit and the School of Clinical Medicine could no longer support our Department. In the current academic job market, I am experiencing some uncertainties about where I will relocate. This is thus one of my last opportunities to show where the discoveries we have done over the last years (in 2022 hopefully you will see much more in print).

Biolabs and 3d printing. Most people know me for my work in microscopy. However, more than half of my group is dedicated to cancer cell biology. This is the 3D picture of one of my wet lab bays. Opposite Suzan’s and Anna’s workplaces, we have our rapid prototyping workshop. Some of you might notice an old Maker Bot Replicator 2. My first introduction to 3D printing. But the most notable printer is the Form3B by Formlabs in our custom (orange) enclosure, with which we print both biocompatible scaffolds and chambers for our light-sheet microscope. Then the Ultimaker, a workhorse for all the needs. The last entry is a CellInk BioX, for 3D bioprinting that we are still integrating into our 3D culture workflow.

Laser lab. Ready for more tech? This is the room that bridges past and future developments. This laser lab was the first lab I could call ‘my lab’, where I started to work in 2010 as an EPSRC LSI fellow on my own. All new tech I produced over the last 12 years came from this room, at least initially. One side of the room serves as our electronics and optical workshop and then we have two optical tables dedicated to prototyping. Certainly a crowded room. The development of our SIM/FLIM system (ATLAS.ONE) was delayed by the pandemics and now we disassembled it getting ready for relocating. Before ATLAS.ONE, this room hosted the various iterations of confocal spectropolarimetry I developed over the years. The centrepiece of this room is now an open-top light-sheet microscope (ATLAS.TWO – CRUK funded). This is just the first preview of a system that hopefully will be the protagonist of many future papers.

Optogenetics lab. After a major infrastructural refurbishment, I was able to get a second laboratory, which I dedicated to optogenetics. Here, we prep samples in a (blue-light or red-light) darkroom. We also have small incubators to ensure keeping cells in light-controlled areas. There up on the shelves, there is the skeleton of our first OptoFarm, a system to culture cells under tightly controlled light (biochemical) protocols. Now, this is discontinued and replaced by a much simpler and more flexible system that we’ll publish soon. And yes, of course, you see also our workhorse for single-cell fluorescence dynamics, integrated with multiple cameras, photoactivation capabilities, multiple light sources, and microfluidics. This commercial system was bought and then modified with an MRF grant and despite being very temperamental, it gave us a lot of good data!

Biophotonics lab. I hear you asking… what about FLIM. Of course, we are almost there. Here, we dive into one of the rooms of the imaging facilities where I customized a multi-photon / confocal microscope with time-resolved technologies. In this front view, you see the Leica SP5 and the Chameleon Vision 2 (to the right). Two instruments that gave me a lot of satisfaction. The blue boxes are a custom-built FLIM system (ELIS) that I built when FLIM was still relatively slow. But now commercial systems are also super-fast and I have packed ELIS for good. To be unpacked, once I will have new laboratories, the PicoQuant rapid FLIM of the latest generation.

Let’s go around the table because this room is full of tech. In this room, I hosted several generations of HDIM systems. Some published, others not. The black box on the table is a streamlined and efficient version of HDIM, a time-resolved spectropolarimeter. Coupled with the multi-photon microscope, we get high efficiency in detecting fluorescence with 16 concurrent spectral channels, 2 polarizations and 64 time-bins. Under the table, is the amazing SPC152 – the heart of the system by Becker&Hickl.

Yes, a lot of boxes around because we need to pack up! The back of the table hosts a pulse-picker we used with SPAD arrays, beam conditioning optics and HDIM Gen 4 (I think!) 🙂 I stopped its development because of COVID but soon or late I will resume. Hopefully, this will be fully automated also in its alignment and will integrate fast FLIM electronics.

This first virtual tour ends here. Hopefully, the next tour – with a bit of luck – will be from my new labs.

On the misuse of case studies: a case study

Our organization is committed to equality, diversity and inclusiveness. For example, Dr Clara Madeup benefitted from our ‘return to work’ programme that permitted her to come back to work after an extended 2 years maternity break. Clara is now a tenure track associate professor leading in the field of biotechnology.

How many Claras and Johns showcase success stories across our industry? More often than ever, we need to submit case studies during assessment processes, so much so that it is not unlikely to receive negative feedback if we describe our actions and outcomes carefully but without illustrating case studies.

Which is the likelihood that an organization does not have good case studies to showcase? And how likely is it that an organization decided to illustrate a failure in a case study? How representative success stories are of an organization, particularly organizations that are based on high staff turnover and competition? In fact, a few handpicked case studies can conceal otherwise worrisome statistics available within a document right alongside nice case studies.

Of course, the exclusive use of positive case studies in our websites, the brochures we use to describe how great we are, or at least we want to be, is absolutely obvious and legitimate.

I have seen case studies related to negative events within my organization only in two cases. First, introductory courses for health&safety that often provides plenty of examples of incidents with few cases discussed in detail. They are very informative because in the utter boredom of a long H&S course they actually tell you the story of not what can go wrong but what did go wrong in a lab like yours, maybe next door. Second, I had volunteered for a course designed to inform how to help victims of rape and sexual harassment. Instead of dwelling on how good our organization is, we went deep in describing which problems we have to deal with, how problematic communication can be, and how both academic and justice systems can easily fail victims. Very different situations but the illustration of what CAN and what DOES go wrong was absolutely instructive and helped focus on what we should do to prevent incidents.

During management meetings, we usually discuss what we can improve. Obviously, we do not speak about positive things only, quite the contrary. However, we do this often through rather unevocative statistics and get excited when we see progress compared to the past, or we are better than other organizations in the same area. I wished, however, organizations would focus more on the investigation of negative case studies during management meetings, of course, anonymized and taking any necessary precautions or even with the consent of colleagues involved, so that we could understand more deeply the consequences of our failures and identify better strategies to eliminate or mitigate our shortcomings.

I think we should bring a bit of the scientific method we experimentalists are so accustomed to deal with. We often learn a lot from experiments that fail for no apparent reason, and we showcase our failures to colleagues to get help and to teach less experienced how to identify solutions.

I am not really sure about how often ‘negative’ case studies are used in academic management to inform executive decisions in the broader community. In my experience not enough, probably, because the ‘negative’ case studies we should analyze are often just simply buried, swiped under the carpet, a topic for more specialist discussions reserved for those that make issues disappear.

I hope organizations will adopt more the use of ‘negative’ case studies as a tool to improve and fully understand the suffering of those who find themself in challenging situations. And I hope we are asked to produce case studies to illustrate success stories and good practices less frequently during an assessment, reserving these to public brochures.

Against (online) abuse

English football has announced a three days boycott of social media to raise awareness against online abuse. I am no footballer and I even do not follow football but I follow Formula One, and thanks to Lewis Hamilton engagement against racism I got aware of this initiative. Sport should be all about coming together in a joyful way and transforming the instinct to compete and fight into a game based on fairness and respect. However, far too often sport – and football in particular being so popular – makes itself a vector of abuse, online or in person, verbal or physical.

Abuse is part of our society and – like all of the human shortcomings – it will be never fully eradicated. However, abuse should still rise indignity from all us, either if online or not. Sadly, influential people have contributed to normalize online abuse, attitudes that are then percolating back into the streets. Admittingly, every person might have a different sensitivity and personal judgment about what ‘abuse’ is beyond the strict legal definition. This should not be, however, used as a free pass even just to be unkind, certainly not to be, well… you guessed… abusive.

Then, as someone active on social media, despite not being an athlete for the past quarter of a century, I’ll turn my social media off until Monday night in solidarity with this initiative and any victim of abuse.

A brief journey to India, and into models of carcinogenesis

In early 2016, I was asked if I wished to speak at the discussion meeting “Conflict and Competition in Cellular Populations” in Bangalore, India organized by Dr Sandeep Krishna and Dr Sunil Laxman (NCBS). The title sounded so intriguing that I accepted without even checking the actual topic of the meeting. Then an adventure begun, that now concluded (did it?) in 2021 with a small paper entitled “Cooperation of partially transformed clones: an invisible force behind the early stages of carcinogenesis” published in the journal of the Royal Society, Open Science (10.1098/rsos.201532). Let me tell you the story of this journey that, perhaps, might inspire you to adventure outside of your field.

For brevity, I’ll skip the details about the actual trip. It was of course exciting to experience a culture I am often exposed but I never lived. The food, the people, the contrasts of India, a small glimpse into a complex galaxy of humanity. My short trip to India started with a sleep-deprived-me trying to explain to the border police that the conference Conflict and Competition in Cellular Populations, nicknamed CCCP, which poster was written in pseudo-Cyrillic, was not a political conference (I would have needed a different visa in that case!) and concluded back in Cambridge a week later with a slightly embellished bedtime story for my 3 years old daughter about the animals I saw in the park that hosts NCBS, a story that I am still telling now and then to her.

But of course, here I focus on the science. The conference hosted a good number of great speakers (referring to others) on the topic of ecology (er, yes, the title made sense). Suddenly it dawned on me I was ‘a bit’ off-topic. However, I loved talk after talk learning a bit about ecology, including its mathematical foundations. I really enjoyed the meeting, so much so I could not stop thinking about its relevance for my work that back then was focused on non-genetic heterogeneity in cell decisions, carcinogenesis and the DNA damage response.

The study of cancer as an ecological problem is not new, of course. Something very specific started to bug me though, something I could not find literature about. We know that different clones of cancer cells cooperate and compete in tumours but what happens during the very early steps of carcinogenesis? I was queuing to board the airplane when I succeeded to download the paper “Evolution of cooperation among tumor cells” published ten years earlier by Axelrod and colleagues in PNAS. It was a nice in-flight read, but the flight from Bangalore to London is long and I started to obsess about a very simple mathematical fact.

For a moment, let’s imagine you dream of establishing a business but you need £1M to start it. However, you are a bit of an odd person and decide to do it only if you win the lottery which jackpot is £500k. You clearly make strange decisions but I am not here to judge… the oddest thing is, however, that you bet on winning the lottery not just once but twice. Then you have an idea You agree with your village of similarly odd-minded people that if anyone wins the lottery, you will pool the money together to invest in this start-up. This is still an unlikely strategy, and certainly one that has a tiny probability to succeed, but it is definitely more likely to work out than waiting to win the jackpot twice alone.

Back to carcinogenesis. Every day, each cell has a certain probability to mutate because of exposure to radiation, chemicals or simply the chance of errors of biochemical machineries. Mutation after mutation in the right genes, a cell might grow into cancer. A very unlikely series of events that, however, with trillion of cells in our bodies, over one’s lifespan is likely to happen. We know that certain mutations occurs in cells that eventually lead to cancer. We know that one cell wins the macabre lottery of disease multiple times before leading to cancer. We then know that many cells will get mutations within an otherwise healthy tissue.

We usually consider that all these other mutant cells will either accrue neutral mutations (i.e., mutations that will not change the fitness of the cell, nor confer a cancer phenotype), or deleterious mutations that will be purged by tumour suppressive mechanisms. However, cells within a tissue communicate and mutations occurs also in genes responsible of cell-to-cell communication. In my recent work I propose a ‘toy model’ with which I explore the possibility that the gene- and cell- centric mutational process should be reconsidered in the context of an overall tissue where cell-to-cell communication might reshape the early steps of carcinogenesis. I am not the first one doing so, but I try to emphasize with simple modelling how the mutational process should be seen in the context of a collective of cells rather than in a gene- or cell- centric fashion.

What did I learn beyond what I have written in the paper (i.e. in addition to the science)?

First I had really fun, something that over time does not happen with every paper, even those more important ones where we invest major resources in. I even had fun during the revision process. As many of us experience, I often got half of the referees very supportive of my work and half rather dismissive. But those very supportive have been often extraordinary kind and helpful, either defining the manuscript ‘a refreshing read different from what I usually read in this field‘ (earlier submission in a different journal) to ‘the models presented here make the point in a clear and dramatic manner‘. The last referee of the last submission now published was particularly helpful. Not only they critically review the manuscript but also invested time to describe a discrete time Markov chain model that I could have integrated in the manuscript. This suggestion permitted me to learn a bit of maths I did not practise before, and to improve the work… this is what refereeing should be.

Second, alongside the enthusiasm of adventuring in a rather different field from my already eclectic research interests, I also felt the pain of being an outsider; a pain I feel often but that it was made sharper by the fact I was a single author. This was really a ‘pet project’. I got convinced to shape my notes in a manuscript only after I attended a seminar by Prof. Allan Balmain in 2018 related to the Nat Cell Biol article “Multicolour lineage tracing reveals clonal dynamics of squamous carcinoma evolution from initiation to metastasis“. It was a great talk and somehow relevant to the notes I had written since my trip to India. I decided to try to publish my ideas after reading the commentary by Prof. Kornelia Polyak and Prof. Michalina Janiszewska where they state: “One possible explanation is that there is a cooperative interaction between the streak and bulk tumour cell populations; an intriguing hypothesis that warrants further investigation but was not tested by Reeves et al.⁵. The streak pattern observed by Reeves et al. is reminiscent of the streaks generated by non-mutualistic budding yeast analysed by Muller et al.¹³.” Eventually, I am not sure the work I had put in this manuscript was worth the pain.

Then, do I advise others to adventure so wildly in other territories? As I have written before, it is rarely rewarding career-wise and never easy. But, once in a while, let’s just follow the passion and enthusiasm for something new, with no regrets. Any adventure comes with some pain but the fun of exploring, eventually, makes the experience worth living overall.

I wish that this small new paper can really provoke some thoughts, or inspire some young scientist to adventure… perhaps not too much and not alone as exploring comes with its perils.

Changing of the Guard

I more excited than other times for a talk I will deliver next week. When invited, I read the list of speakers and I noticed so many names of people whose science I follow very closely. This time something is different though. I read their papers since I am a student, papers they published perhaps when they were students or young postdocs, in fact many of them are my generation. I grew with their science as if they were well-established academics as I never paid attention to affiliations or titles. Some of those I had recognised early in my career disappeared from the field or academia, others are fully established by now. This made me thing about my attitude towards generational change… a great contradiction of thoughts.

Missing the Old Guard. Several scientists I really respect have retired or are about to. I have been privileged to meet so many, particularly in the area of biochemical/biophysical imaging. Scientists who contributed so much, inspiring figures who shaped contemporary science, often without hype or even recognition in the broader community. Wait, am I missing the Old Guard? This feeling contrast so much with another one. In time, old ideas become an obstacle to progress and a generational change is desirable. You might indeed know the popular concept that ‘science advance one funeral at a time’. I do think there is an element of truth in it. So, why do I have such profound contradiction in my feelings?

Loving the New Guard. I am active in the area of biophotonics since an undergrad student, and having swapped discipline a few times, it is simpler for me to use microscopy as an example. The super-resolution revolution has been inspirational although I have observed it from the outside. In a few years, a new generation of stars begun to shine and a constellation of younger scientists who broke with the past was born alongside. Also in biochemical imaging I see great changes, the consolidation of certain ideas that once were considered heresy or simply very very niche. And yes, this get me rather excited. Wait, do I really love the New Guard? I see so much I do not like in science, and this is not just something imposed or inherited by previous generations. There are so many colleagues* with whom I might disagree about science and often on how Academia should be run. Disagreement is ok but sometime this is a much more profound divide.

OK, I got it wrong. Today, I have suddenly realised how wrong I was in interpreting my own feelings about generational change in Academia. While the majority of us would agree that generational change is necessary to avoid science stagnating, perhaps we do not really understand why**.

I love challenging established ideas on the basis of logic and experiment, I love discussing alternative interpretations that are not mainstream (but still scientific!), I love risk-taking in science (not in life although sometimes it is difficult to keep them separate), I love intellectual change (not so much change in my everyday life). Generational change might help the things I like to emerge but old generations do not have exclusivity in being dogmatic or risk-adverse, indeed those I admire are not. The issue is that too often also the younger generations accept dogmas (not just critically incorporating established theories and models in their thinking), they would guard an old ‘truth’ no matter what. But when they lose their authority of reference because of generational change, somehow their confidence or power is weakened, leaving space for positive change.

Hence, I now realise I am merely recognising a new generation of scientists with whom I might share a vision and I am excited that new people now replace those who retire for whom I had the same affinity and respect. Generational divides are much less important than an open attitude to change.

So, perhaps, I do not like guards in science at all because in science the fewer cages or palaces we have the better it is.

And after this lucubration, I will thoroughly enjoy my next talk in any case 🙂

NOTES

* I use the term colleague very loosely to refer scientists in related fields.

** I just had a glance to this paper by Azoulay et al., interesting concepts

COVID | data analysis (new CAAT 4.3 release)

WARNING. I am not a medical doctor nor an epidemiologist. The analysis I am sharing here is only for the data geeks around that are curious. Please follow the advice of your national authorities and health system.

I have just published a new release of CAAT, Matlab code to analyse Johns Hopkins dataset on the SARS-CoV-2 pandemics. The usual caveat is that data is likely to be underestimates. Underestimation does not occur only because of lack of transparency but most of the times because of differences in definitions of COVID-related fatalities and efficiency of reporting systems across countries. For example, using the excess mortality statistics we know that about 30-50% under reporting is rather physiological because of deaths occurring outside the hospital settings or because people might die positive to COVID but not for COVID. This also accounted to a significant adjustment of statistics we have noticed in the past in Hubei. However, changes over time within a country might be more reliable and, therefore, there is still something to learn from this data.

I started analyzing data when the UK government decided to drop the policies set up for containment of SARS-CoV-2. Now that data has been extensively discussed in the UK, I curate CAAT only for others that would like to explore the data. However, at this release it is worth mentioning just the main observation. While the first acute phase of the pandemics is subsiding in Europe and Northern America, it is now flaring in South America at worrying pace.

Relative fatalities in the population at risk for each country. This is a new visualization I provide. Both bar graphs are ordered according to the values shown to the right, i.e. the fraction of the population at risk who has already died. At the right, the weekly rate of fatalities during the current week and the preceding two weeks. Those countries that, so far, has experienced high casualties are showing a significant reduction of weekly fatalities showing the positive effects of policies aimed to contain the virus.

Commenting out line 362 ‘d_ord = p_ord;’, you can order the weekly rate of fatalities according to the last week. This might reveal which are the countries at most risk now. Several South American countries are topping this list. In Brazil, Ecuador and Peru, about 0.5% of the population at risk died during the last week and, sadly is several countries this rate is accelerating.

Let’s also focus on some good news. This is one of the several plots hard coded in CAAT, comparing Italy, UK, Germany, Denmark and Sweden. As it is well established by now, new fatalities are dropping significantly and lock-down measures are gradually abandoned in favor of social distancing measures. The individual outbreaks started at different times and at different rates. Comparisons between countries are therefore difficult but the effectiveness of policies within countries can be evaluated. If you were interested in different graphs or comments, let me know but I will not elaborate more on this at this stage.

This graph is a bit of a mess but I present it for completeness. In North America lock-down measures are having a clear effect. In South America we see alarming increases. As I do not follow South American politics and specific news, I can’t draw conclusions, but it is evident that – assuming no change in reporting occurred, Argentina has slowed down the outbreak but not yet put it under control.

Although I have explained this before, I should probably clarify how I evaluate the population at risk. To estimate how many people might die in each country in the (unrealistic) scenario where everyone would get ill, I used the age-dependent fatality rates published in The Lancet by Ferguson’s group and multiplied these values with the demographics of each country as reported by the UN.