Welcome to a new episode of the STEMterviews. Today, it’s with great pleasure that we have with us Elisabeth Bik. I’m a huge fan of her work as well. I follow her on Twitter for a while now, I always follow her little image puzzles as well, which is a lot of fun. So we’re really happy to have her today with us. Hello, Elizabeth. Hi, how are you? Alright, thank you. So to introduce yourself, would you mind talking about your latest microbiology research project that you worked on? Well, my latest project I worked on that’s a while ago, actually, I don’t work in microbiology anymore. But I used to work at Stanford and I worked on a project that involves dolphins and looking at which microbiome they carried. So which bacteria do they have in their mouths, in their stomachs, in their rectal area, and their blowhole. Things like that. So that was a really awesome project to work on. because it involves working with dolphins and you know, what’s not to love? Yeah, it is I wasn’t able to, you know, sample the dolphins myself, that was all done by hand, you know, veterinarians and trainers. But I got a glimpse of which bacteria they carried with them. And so they carried very different bacteria than most other mammals. And so they’re very, yeah, very unique in that respect. But they also, of course, they swim in the sea. So we were also wondering if they would maybe pick up some of the bacteria that you would find in seawater. But that was not the case. So they’re very different than so somehow, they selected a very different group of bacteria that they carry with them. So that was one really cool project. And I, I also worked at two microbiome startup companies. One is currently being investigated, because the owners are, the founders are charged with fraud, but it was still an interesting project to work on. I was the science writer. And so we, I made sure that the signs of the project was okay. And what we did is we had a project, looking at which bacteria you carry. So it was for, for consumers who would who are interested in which bacteria they carry with them. So we offer to kids to look at that. Oh, well. So why why is the project important? Why do we need to know which bacteria are in dolphins? What do you learn? A lot for the dolphin for the people who manage the dolphins, it was important because sometimes dolphins get sick. And so because there was so little known about which bacteria they normally had with them and they were healthy, then it makes it harder to recognize what makes them sick. So if you have a sick dolphin, and you look at, you know, and you find particular bacteria, you you need to know which bacteria they normally should carry with them. And so there was just nothing known about that when we started this project. Or there were like, like a couple of a couple of papers were had been done. But most of these were culture based. And so we had just done surveys of the human microbiome using molecular techniques. And so this was relatively new, about 10 years ago. And we were asked if we could do that for dolphins. And so we made sort of this catalogue of which bacteria they have been they were healthy. Oh, okay, so what what would be a typical dolphin disease caused by bacteria, or microbes? They have. I don’t know too much about that. But like they can have gastritis. So So some of the problems that so this was a group of dolphins that were managed by the US Navy, they were sort of like, similar to sniffing dogs, they were trained to find underwater objects, and some of the problems that they occasionally had where it was stomach problems. So like, stomach inflammation, and so they weren’t quite sure what was causing that with humans that’s often caused by Helicobacter pylori. That’s a bacterium that causes stomach disease in humans. And we also found some helicobacter in these dolphins, but we also found them in healthy dolphins. So we were not able to solve that puzzle yet, but I’m sure it’s the basis of new research, like any research builds upon previous research. Yeah. Okay, that’s interesting. So somehow the dolphins are resistant to the helicobacter bacteria, doesn’t mean we weren’t sure we weren’t sure so so maybe they were like, about to get sick or so we weren’t sure, but some of them carried back to with them, but the numbers were just too small. We had like one or two sick dolphins and then a couple of healthy dolphins in which we found the Helicobacter there. So we were in and we also found different species, like in humans, you only have Helicobacter pylori, which is the bacterium that, you know, causes classic ulcer disease. And, but but in these dolphins, we found different kinds. So we weren’t quite sure if there was one particular species of Helicobacter that was causing the disease. And if the other ones were just innocent bystanders, but the numbers of sick and healthy dolphins we had at that time was not large enough to really do any statistics on it. That’s cool. Nice. And have you done any microbiome studies in humans as well? Yes. I was involved in one of the first surveys of the mucosal microbiome. So this was in, we published it in 2006. So that was one of the very early microbiome studies, one of the early molecular look at that which bacteria people carry, and we back them, we did three patients, we had biopsies from different parts of their intestinal tract, so mainly the large intestine. And we also had stool samples of these people. And we did a inventory of which bacteria they carried, and it was only 3 patients, nowadays, some studies involve 3000 patients in that time, that was a huge number, and then the sequencing and data analysis was also not computerized back then. So it was all like manual alignments with a lot of work. Yes. Okay. So what was your most exciting finding back then? Um, the finding was that the mucosal so there, the bacteria that were that we found on biopsies were very similar across the large intestine. And so there wasn’t a lot of variation. And also the stool sample could be easily matched to the donor. So so if you compare the mucosal and the stool sample, they were a little bit different. But they they could you could clearly match it. So we sort of concluded back at that time, that the mucosal you know, the adherent microbiome might be very similar to what comes out. And that stool samples could be a relatively good way to sample the what, what is adherent to the mucosa, of the intestinal intestinal lining, basically, I think that view is now has changed a little bit. So there are a little bit of differences, like some, you know, the bacteria that pass through our label different than the bacteria that adhere to the mucosa. But, you know, it’s so much easier to get a stool sample from a patient or a healthy person than to have them come in to colonoscopy take biopsies. I mean, that’s a big inconvenience for the patients. It’s not without risk either. So we never know when he made a contaminated somehow, either. Also, also, yeah, but it’s just you know, a hassle. If you ask people Oh, do you want to donate the stool sample that is just so much easier to go in for a colonoscopy. So it’s, we sort of concluded that stool samples were a decent type of sample to look at which bacteria live in your intestinal tract. Nice, nice. So now you’re not in academia anymore. You don’t do active research, anybody sometimes miss it. I miss lab work. But I’m also getting little bit older. So all these little tiny letters, you know, I used to laugh, read a book that I reached that age. And also, you know, lab work involves a lot of physical work, like crawling on the tables and moving heavy equipment, and it was almost a little bit harder to get older to be quite frankly, but I, I really enjoyed lab work. I still feel very much involved with science, even though I don’t really do like wet lab work. I feel I still do research because I’m looking at a lot of different papers. So I still feel very connected to that. Yeah, yeah, I feel the same way. Even though you’re not doing active research anymore. I still feel like a scientist because reading all the science papers, you still think about all the experiments and everything. Yeah, it’s still it’s really exciting to be a part of it somehow that’s at least right? So you kind of want to switch a bit to your now the work you’re doing now. You left academia to become a full time science consultant, and you help improve the scientific integrity of science. We’re not going to talk too much about how you got there because you’ve told that story many times that I just wanted to talk about like, you gain a lot of attention for the work you’re doing now. Like there’s a lot of trolls going on on Twitter, like even now when we announced our little interview here on Twitter, there were a few responses which I was like what why is that happening? Where do they come from? How do you handle this kind of rejection because you really believe in what you do and it’s so important but it must be daunting as daunting as well to sometimes always get like all these critics critics. How do you deal with it? Yeah, it’s it’s tough. I’m not gonna deny that but but you sort of grow into it. So the first negative comments is, you know, that hurts a lot then and some of these still hurt, they get harder and harder. So Yeah, it’s it’s never fun. But I also have to keep in mind that the reason I’m getting them is because I have ruffled some feathers I’ve, I’ve created some Yeah, some anxiety, maybe with some of the researchers whose work I found errors in or maybe even suspected fraud. And so I know that some of these researchers might have groups of trolls that they either pay or somehow encouraged to harass me. And so I yeah. I never thought of, I do think I do think that some of these trolls are like bots or like groups of people who you know, you can buy followers on Twitter, and some people have incredible amount of followers, even though they, if you look at their followers, some of these people never tweets or they they all tweet the same thing at the same moment. Like Like, suddenly there’s these outbursts, and so it appears to be coordinated in some way. And so that’s not pretty bad it is. But I do feel that some of these groups are very different from the followers I have who seem to be real scientists and real people and you know, people have life stories and pictures of cats. And these trolls don’t don’t have that, like they, all they do is negative comments. And maybe they’re very frustrated people or I just picture that somebody is paying them for to do that. So crazy. Yeah, not sure how it works. So maybe I’m completely wrong. But yeah, they don’t know what what fun is it to just have negative comments about people and rescue me away. But ya know, it’s it’s, it’s tough, and it’s but I do try to focus on the enormous amount of supporters I have, and knowing that I do the right thing and knowing that these researchers who have negative comments about me don’t have answers to the research questions. I’m asking them so I have concerns I post these old ones, and they don’t answer those questions. And since that they they attack me that attack my work, my appearance, whatever I say, all my my previous paid for so everything is ripped apart every announcement where I have an interview, they will they will be there to follow my every move. But it’s because they don’t the researchers themselves don’t have answers to these questions. So it tells me I’m I’m on the right path. I’ve asked the right questions. I guess so yeah, yeah. And how do you deal with like being in the spotlight Like this? Like, do you protect your family? Or do you try not to disclose too many private comments or private information? How do you do that? Yes, I am a bit more careful about posting where I am. I also hope that at some point, these harassment attacks will will stop. But yeah, one of the trolls are actually one of the professor who tweets under his full names in France who, who has been very faithful in attacking me. He has published my home address, and that is that is of concern. I mean, it’s, it’s sort of you can find it sure. But not to post it that just tells me that he is, is asking people to maybe attack me, personally. And that that is of concern. Yeah, definitely. I’m sorry to hear about this. Because you have all these like negative comments on social media. And it’s obviously really tough what you’re going through, how do you think the scientific community can support you and your work and help protect you, so anything you think people can do? Well, I already have received a lot of support or two petitions to support me. And that has made such a an enormous difference for me, because it did feel very lonely being attacked by a lot of trolls seeing all these negative comments that does eat on me. And so once these two petitions started, I just knew, okay, what I do is the right thing, I should look focus on the positivity, focus on the support. And that has just made an enormous difference for me personally. But on a on a general, more general level. I guess another thing that I hope the scientific community would be would be doing is to better screen papers before they’re being accepted. For things like fraud to be more aware of fraud, and for editors. So for journals to be quicker in their responses to be more open for concerns that people like me are sending to them. And to retract papers. If there’s big concerns in a paper. Don’t accept a new set of figures or don’t accept, you know, an excuse like oh, you know, here’s a new set of data points are so and I will just do a correction. No, that those papers were very, very suspected that there was fraud. Those should be retracted that should send a more clear signal to the researchers who are doing this I I feel there are a lot of the editors are too gullible. They’re too nice. They’re Yeah, they’re not open for the idea that some of these things might be fraud. And we’ve there were a couple of recent cases and very fast retractions, and I hope I have like 4000 papers reported and only a fraction of these have been retracted. And it’s just a shame how long that takes. That’s, that’s incredible. It’s so much work. And yeah, if nothing happens, it just like often nothing. Okay, so is there any one app papers that does kind of the same work as you do like us? Yeah, looking at the quality of images or anything? Or do you think this should be something that it should be implemented? Just as editing the text of a papers implemented by an editor at the journal? I should maybe also some one, yeah, quality manager or whatever? papers? What do you think? Well, yeah, at the journal level, there are new developments. So similar to when manuscripts are screened for plagiarize text. There are now people working on software to screen the images for plagiarism. So with that, I mean, like duplications within an image duplications across different figures. But also, this is something a human could never do, comparing one particular photo to all photos that have been included in all scientific papers. And that is, of course, computationally, enormously challenging. And currently, that’s not possible, but as to the best of my knowledge, but I do feel this is something that will be solved in time, like when computers get more stronger, they can, they can easily do that. And so that opens up, I bet they’re gonna find a lot of images that have been reused. It’s just super hard to recognize, I sometimes try with a reverse image search on Google. And occasionally I’ve been lucky but in most of the cases, Google image search is not designed to find one particular panel that has been taken from another paper it will just look at the whole figure, you know, with all the different panels and it takes that as it’s search doesn’t look for particular smaller components. So it doesn’t really cannot detect these things, but I bet there’s a lot of reuse with images. So are you actually involved in developing these kind of software’s. Has anyone approached you with this? I’ve been approached plenty of times and I’ve shared a dataset so I made a dataset of 300 something beyond the 50 papers, all from the same journal in which I had found problems of duplications within that paper and I made a matched control set of papers that were published in the same same issues of those journals of the same journal and around the same time so it was a good set in which I did not find problems so you know, maybe there were problems but at least I didn’t find them so it’s not a perfect set, but the best I could make and I’ve shared the data set with I’m gonna say at least 20 different groups who claim together that they could work on software to catch these things. But most of these groups have never called me back or provided any update and I know there’s now some groups that have have prototypes but I have not been offered to work with any of these so it is frustrating because I share my knowledge Yeah, but I’ve not been involved with the more developmental stages or with the commercial aspect because of course the software could be bought by publishers and could earn some money. So that’s been frustrated. On the other hand, there are a couple of people who who have shared so I’m actually beta testing one of one of these programs I never shared my set with with them but but this is image twin an image twin is is working really well so I have actually a access to this software sort of still in beta and still being developed. But it works really well and it helps me a lot because I guess I could find these duplications myself but the software is so much faster Yeah. It also fails to fail so miserably sometimes sometimes I clearly see the duplication and the software cannot find it but but it has helped me a lot in these you know in particular where you have these very complex panels with like I don’t know 20 different panels that all look very similar to each other or tissue sections or so the software is just very fast it just within seconds it finds the duplications and that would maybe take me hours so that saves me a lot of time. So I using this software for your Twitter puzzles as well. Sometimes Yes. What do you find and what people find or what does the software find? Well, I would have scanned them I’m already with the software before I post them online just to make sure that the software also finds it. And yeah, occasionally, the software will find much more than I will find that sometimes he’s very good at picking up tiny little slivers, and is also very good at finding rotations. That is very hard. It’s difficult to see. The rotations are very hard. When it’s upside down. It’s it still works when it’s when it’s like rotated halfway or so like, like just 30 degrees or so I find those are very hard to pick up with the human eye. Yeah. Okay. And just as a takeaway for actually scientists will hopefully watch this. Can you just summarize what is actually allowed in image manipulation in science? What can you get away with? Where’s the borderline? When does the scientists actually cross the line? What right so so most journals now we’ll have more strict guidelines for what is allowed, so exactly this question. And so you are allowed, if you have a photo, or or any other image, you are allowed to change the contrast of the whole image, but you’re not allowed to change the contrast of part of the image. So for example, if you have an agarose gel with a band, and it’s very faint, you can, you know, crank up the whole intensity a little bit, but you kind of just look at that band and crank it up and not crank up any other bands. You know, that’s not a lot, because you could make the relative intensity of the bands. If you have multiple bands, you could you could play with that. So that’s not allowed. You’re also not allowed to take one cell in a photo and just stamp it 10 times that would be severe photoshopping. That is, I mean, I think most people will will agree with me that that is not a good thing. Like you cannot say, I see 10 cells, when it’s just 10 copies of the same cell that’s definitely not allowed. So you are allowed to do like things overall over the whole image, but not a part of the image. There’s more particular rules. For example, for western blots where you’re should have, there should not be splicing across lanes, so you cannot merge two different western blots with each other. Your if you do that, you have to make sure that you show that there’s a line in between. That is something that is different, because when I was doing my PhD research, I actually have a paper in which I have to splice, it’s not a duplication, and that was totally accepted that you just had, you know, a gel with 12 lanes, but you had 13 samples. And so you just had to, you know, run it on another gel, and you just literally cut out and glued it together. And that was your photo. And that was totally fine. As long as you made sure that the markers were lined up at the same length. And I, I did that. So I do have a paper in which I have a splice that would not be allowed nowadays. So that’s sort of change of insights. And there’s also rules, if you if you just show a couple of bands, that you have to have some height, and above and below the band so that you can just like just cut out the bands and you barely see the environment. Yeah, we had this antibody that would always recognize something just above my like my band of interest. And my bosses just cut it out. Well now it would be like, should we do this? Exactly. So it’s just disturbing, because there’s two really strong bands, but only one is of interest for us. So yeah, yeah, that there are new insights and these rules. And I think most scientists, you know, they do these things. And they do it very honestly, like you just said, like, you just don’t want to show that band. But I it’s also been misused. And so the rules have to be stricter, because there are a lot of fake papers nowadays that being are being generated. And some of these are with computer generated AI generated images, western blots that are completely not real. There was never a real western blots. It’s all computer generated. And that is a very scary development, because there’s no duplications for me to detect. The bands can look incredibly real. But yeah, there’s just no background that is that looks like a normal thing. But of course, backgrounds can also be generated. What do you think it’s the motivation behind this just to have the perfect Western Blot? Or what is the I don’t understand? No, it’s so there’s this whole sort of organized crime almost like where, where people need papers, and people provide papers that are completely fake and people make money selling papers, selling authorships, to people who need papers. And this is a very particular problem in China where there’s a particular requirements for medical doctors who want to have a position at a not at a research institute, but at a clinical hospital. They need to publish a paper. And they’re not actually interested in research or interested in doing clinical work. But they still need to have a paper published. And so sometimes they just buy papers. And there are we call them “paper mills”. So these are sort of companies that sell papers that we believe are completely fake are just written based on a template and written. Yeah, by ghost riders, and they contain fake images, they contain fake data, it’s just not real. And those papers are completely fake. And they’re hard to recognize, because they’re unique enough to fly under the plagiarism detection radar. But they’re Yeah, they’re very. They’re polluting basically the literature because those experiments never happened. But that is crazy. I mean, there there must still be based on some data somehow. I mean, they can’t just make up data, or do they that was happening, they just take a random topic and just make up something else that they make. Well, they’re they’re all based sort of in the same, they’re all in the same topic. So they’re all like non-coding RNAs and cancer. But they replace the name of one non-coding RNA with another one, they replace one cancer with another cancer. So there’s papers that all look very similar, if you compare them next to each other, you sort of start to recognize the templates. But one is about prostate cancer. The next one is about gastric cancer, and they make up a fake new set of patients. And sometimes they have made mistakes, or we have a bunch of papers in which if you look at the patient, characteristics of the there’s a table in there, but the patient’s characteristics, in some of the prostate cancer papers, you’ll have half of the patients being female being women, and that’s very unlikely if you have a set of prostate cancer. And similarly, there are breast cancer papers that half of them are male, and so men can get breast cancer, but you will not expect half of your patient population to be male. And so they’re in with those mistakes, you can see that the role based on a template and they just forgot to adapt the patient characteristics on the type of cancer. But even that the journal should pick up, no? I mean, even an editor should see what how you would think if a prostate cancer peer reviewer, that that he or she would notice that there’s something wrong with the patient’s there. But no, these are all these old past peer reviewed. It’s crazy. That is absolutely crazy to think about Yeah. Okay. So thank you very much for your time for talking about all of this has been really interesting. So at the end, we always have a couple of random questions, which we like to random questions. Okay, interviewers, and it just that they don’t have too much to do with the actual work is just to get more about you, the person of you. So, should we just start So what was your favorite subject at school? Oh, biology for sure. I love biology. Yes. So yeah, it fits what I’m doing. doing currently. Yeah. And one sentence, what are you truly passionate about? About science being about finding the truth? finding the truth? That’s good. Right? That’s what science should be about. It should not be about making up results. Yes. What do you do in your free time, making puzzles for the Twitter? Yes. Don’t have too much free time. I I love gardening and then the type of gardening that involves, you know, sprinkler repair and like pruning big bushes. I’m very tiny. You know, cutting one leaf type of person. It’s a that always involves me in some big fight with with a big bush or a big tree or a broken sprinkler. I love doing that. It’s good. What is your favorite movie including bacteria? I don’t know any movie with bacteria. But I know “Contagion” is about a virus. I don’t know if that counts. That’s, I’ve only seen it one contagion. But basically, we’re living in contagion right now in the sort of same topic as that movie. And I was very impressed with that movie. I also thought it had some big mistakes about, you know, representing how science does not work. But yeah, overall, I think that was a very scary movie. And you know, it all came pretty much it all came out. So yeah, it’s not about bacteria, though. It’s difficult to watch those kind of movies these days. I remember a couple of days ago, a friend told me Oh, you should watch this meet new Netflix series, which is called “Bio hackers”. And I just started and it was about synthetic biology and they did like PCRs within 10 minutes in the lab and I was like, No, I cannot watch this. What is this? Why? No, no, no. Yeah, it’s good. Okay, and our last question, which is actually my favorite, what would you do if you were donated $10 million to your project? Oh, wow, that’s a lot of money. Well, I would think I would donate a lot to people who who provide the platforms that I’m working with. So I’m working a lot with PubPeer, which is a website onto which I post my concerns. And that’s all run by volunteers, with very little money, there’s Retraction Watch who, it’s a blog about retractions. And they have a data set database set up to keep up with retractions and make them searchable. So I would donate a lot of money there because I feel that’s an incredibly important work. And I would probably have legal funds for people, not just myself, but also for people who would be who try to raise concerns about scientific papers and who are then being threatened with lawsuits like I feel we need. We need a big pot of money to defend ourselves like I don’t wouldn’t have money to defend myself. And I feel that if we if we could have a lawyer helping us deal with these, these nasty cases, that would be incredible. Yeah. Yeah, that would be a good cause. I think with lawyers $10 million would not last very long. Yeah, and it just like infrastructure, like, like, how can we better make warn people that papers have problems? How can we have people retract faster? I I’m not quite sure if that that will take money. But if that if there’s if money can solve some of those issues, that would be incredible. Right? It’s all about also about education on scientific integrity, like offering courses or whatever it might be. Yeah, right. Well, there’s there’s lots of courses on scientific integrity, I feel they do very little in actually addressing the problem. But we do need is like, yeah, and maybe that’s part of the money could go to they’re like, we need an international organization that actually can investigate these cases, investigate, who did you know, the potential misconduct and punish have some punishment? Yeah, you know, I don’t mean people looking up in jail. But like, there needs to be some repercussions for scientists who keep on frauding. Like, and instead, we still give them money for grants. But there should be, you know, some of that money should be taken away from them. They should be banned from getting grants, or they should lose their job because it’s, you know, they’re not doing their job well. And if there’s no repercussions, people keep on doing this, you know, there wasn’t a chance that you could get a ticket when you run a red light. And then exactly, if you would never get a ticket, we would all run the red lights and we would all be speeding on the freeway. It’s only the thought that there might be a police officer behind the bush, giving us a ticket that keeps us you know, driving a safe, safe speed. So we need to have more more, you know, the alternative of the tickets. Yeah, true. I agree. Thank you so much Elisabeth, for your time, and for this very interesting interview and a great conversation. Thank you, so much Sarah. I learned a lot, really great. Thank you. Thank you for watching with STEMcognito. Find more videos using the search box or the drop down menus above. If you think there’s something wrong with this video, please use the Report button to inform the STEMcognito team. Questions about the video content should be directed to the researcher. You can find their details below. Go to our submission pages to find out how to submit your own video and don’t forget to follow us on social media.