Developing scientific reputation relies on publishing impactful discoveries, developing widely used algorithms, but probably equally importantly delivering insightful scientific talks. In some conferences, I am often chagrined or amused by talks in the following styles (sometimes a talk could combine the 3 styles which is quite a sight to behold :):
A speaker goes over all the major publications (Nature, Science, New England JM, Lancet, you name it) to demonstrate that a new direction (often following a US or EU initiative) is important, although none were their own study. It is like a 科普 journal club which treats the other scientists in the audiences like school children. Actually such talks are often good for student audiences who may not be as well informed about the new trend, so are quite informative and inspiring. Also such talks often signal that the government will be spending major $ on this new direction, so we should start to scheme and get a slice of the pie :).
A speaker says “the computational method is very complicated, so I won’t go into the details, but trust me, here is our finding”. Then a list of biological observations is given, and the speaker says “we don’t know what this means” or “I don’t really understand the biology”. It is possible that these methods are good and the discoveries are true. However, for these results to be taking seriously, the data / method need to be explained to earn credibility, computational biologists should learn the domain biology and talk more professionally about the biological findings.
A speaker says, “the method is very simple la! so I won’t go over the details, but we have boat loads of money to generate tons of data, and we integrated everything, so here is our results”. Then to the abhorrence of the audiences, the “results” presented are “we filed a patent or developed a kit”, “we obtained some major government awards or venture capital funding”, “this method can really be used to solve all biomedical questions, which can be demonstrated by the following 100 published papers from our group”, or “we found a new use of an old drug or a new drug combination, some have clinical trials ongoing, but for intellectual property reasons I can’t tell you what it is”. Such talks are often super effective ;), since naive audiences often fall head over heels to join the speaker’s lab.
A good scientific talk should focus on one coherent biological study, start from the motivation, explain what data is used, how the computational method works (at least the key intuition), then present specific biological findings with thoughtful biological interpretations and full appreciation of potential limitations. If the talk is a computational method, then explain why and how it is developed, what the main functional features and advantages are, where the method can be downloaded for use. The audiences could learn how a study progresses from a logical flow, how to a use tool / website / database to speed up their analysis and hypothesis generation, how to use the thinking of the presented method to develop their own methods, how to use the presented biological findings to inform their next experiments. It is OK to present published studies, but probably even better studies under preparation, review, or revision (many reputable international conferences encourage speakers to present unpublished work).
I hope that with more and more great scientists getting into computational biology, the field becomes more mature, our scientists could learn to deliver professional, more informative and insightful scientific talks in scientific conferences. At least I hope speakers at the Young Bioinformatics PI Workshop try their best to give good talks. Only when we can do so can computational biologists be respected by experimental biologists, can Chinese scientists earn international reputation.
P.S. Recently Xiaowei Zhuang gave a talk at our CFCE retreat. She talked about many different studies and papers from her group, but was super clear on the intuition as well as the significance behind each study. You could see how their technology progresses over time, and could really learn a lot which was very exciting. I guess it is possible to cover many topics in one talk and still do a great job, but Xiaowei’s scientific expertise as well as her talk style (which I am sure take a lot of hard work to hone) is quite a different league!