European Bioinformatics Institute

Scientists are predicting individual risks for 20 different types of cancer with a high degree of accuracy. The prediction model could help to identify people with a high risk of cancer, for whom individualized early detection programs could be tested in studies. Scientists from the German Cancer Research Center (DKFZ) and the European Bioinformatics Institute EMBL-EBI, Hinxton, UK, are using the Danish health registers to predict individual risks for 20 different types of cancer with a high degree of accuracy. The prediction model can also be transferred to other healthcare systems. It could help to identify people with a high risk of cancer, for whom individualized early detection programs could be tested in studies. If cancers are detected early, the chances of cure are generally higher and patients require less intensive treatment. However, screening programs for the early detection of cancer only exist for a few tumor diseases -- and not all people take advantage of these offers. If there were a simple way to filter out people with a very high risk of developing cancer, screening programs could be developed specifically for those at risk. Researchers led by Moritz Gerstung from the DKFZ and the European Bioinformatics Institute EMBL-EBI, Hinxton, UK, have now published a feasibility study on this topic. The data scientists used the comprehensive data from the Danish health register, in which all clinical diagnoses of the population are stored, to quantify the individual disease risks for 20 different types of cancer. The researchers first trained a prediction model on the data of 6.7 million adult Danes from 1995 to 2014. The training data set included more than 1,000 different previous diagnoses, as well as cancer in family members, age and -- where available -- basic body data and risk factors such as tobacco consumption or obesity. The model was then validated on the data sets from 2015 to 2018, covering 4.7 million Danes, and delivered a high level of predictive accuracy. The model enables predictions about the individual risks of developing 20 different types of cancers. Over the course of a lifetime, the model achieved an accuracy of 81 percent. Taking age and gender effects into account, the accuracy was 59 percent. The model achieved the highest accuracy for cancers of the digestive system, as well as for thyroid, kidney and uterine cancer. In order to check whether this predictive performance was also confirmed in health data from other countries, the researchers also validated their model using data from the UK Biobank and achieved comparable accuracy. The analyses do not allow an exact prediction of which person will develop cancer. However, they do determine the individual risk and enable a comparison with people of a similar age. "With this study, we wanted to demonstrate that it is essentially possible to model individual cancer risks on the basis of national health data," explains Moritz Gerstung. Such risk stratification could help to offer further early detection tests to those people who would benefit most. In addition to established early detection methods, these could in future include blood-based cancer tests, for example, which are the subject of intensive research worldwide and are already being tested in clinical trials in some cases. The underlying hope is that in future, a certain number of tests could detect more cancers following risk stratification, with people at low risk could be avoided unnecessary tests and false-positive results and overdiagnosis could be prevented. However, as Moritz Gerstung says, a reliable database is essential for this. "The Danish health data is unique because it covers a large period of time and can be linked with each other. Only a few European countries offer something equivalent, such as Finland and Sweden or special research cohorts in the UK. Efforts are also underway in Germany to establish national digital health infrastructures. "It would make sense to consider which type of data is best suited for assessing cancer risk at the planning stage," says Gerstung. In his current work, the ICD-10 diagnosis codes, which are also used in other European healthcare systems, have proven to be useful. Since basic information on body measurements and known risk factors such as tobacco consumption also provided important information, it seems advisable to facilitate the collection of such data at population level. "If this data had been available across the board in the Danish health registers, our prediction model would probably have been even more accurate," summarizes Gerstung.

All modern Plasmodium falciparum, the deadliest malaria parasite in humans, are descendants of one initial infection and so are very closely related, with relatively limited genetic differences. A long-standing mystery in the field has revolved around a very few locations in the P. falciparum genome where there are 'spikes' of mutations -- far more than anywhere else. Researchers have identified two genes in which these unusual mutation spikes result from DNA being copied and pasted from one gene to another. By dissecting the genetic diversity of the most deadly human malaria parasite -- Plasmodium falciparum -- researchers at EMBL's European Bioinformatics Institute (EMBL-EBI) have identified a mechanism of 'copy-paste' genetics that increases the genetic diversity of the parasite at accelerated time scales. This helps solve a long-standing mystery regarding why the parasite displays hotspots of genetic diversity in an otherwise unremarkable genetic landscape. Malaria is most commonly transmitted through the bites of female Anopheles mosquitoes infected with P. falciparum. The latest world malaria report states that in 2022, there were an estimated 249 million malaria cases and over 600,000 malaria deaths across the globe. 94% of malaria cases and 95% of malaria deaths are found in Africa, with infants, pregnant women, travellers, and people with HIV/AIDS being at higher risk. The new study, published in the journal PLOS Biology, provides key insights into the evolutionary history of P. falciparum through the analysis of two genes that encode surface proteins critical to immune evasion. The genes in question are DBLMSP and DBLMSP2. These findings deepen our understanding of how the malaria parasite has evolved and could help to inform new approaches to vaccine development, offering hope for more effective prevention methods against a disease that continues to impact millions globally. Copy-paste genetics Usually, the sequence of an individual's gene is inherited from their parents, but in some circumstances, part of a gene sequence can be copied between different genes on the same DNA molecule -- this is known as non-allelic gene conversion. This process has been linked to the evolution of important gene families, including those involved in the functioning of the human immune system. One of this study's key discoveries is that gene conversion takes place between the P. falciparum DBLMSP and DBLMSP2 genes and results in increased genetic diversity within the surface proteins of the parasite. Since these proteins are exposed to, and interact with our immune system, they are potential vaccine targets, and a fuller understanding of their genetic diversity could be very valuable for vaccine design. "The discovery of 'copy-paste' genetics within malaria's DNA reveals the impact of an underestimated evolutionary mechanism," said Brice Letcher, Postdoctoral Researcher at the Laboratory for Biology and Modelling of the Cell (LBMC, France) and former PhD student at EMBL-EBI. "Here we show that gene conversion was a potentially important strategy behind malaria's ability to adapt and thrive in humans, including possibly to evade the human immune system. Understanding this genetic flexibility offers new perspectives on malaria's persistence in and adaptation to the human host." Mapping hidden genetic diversity in malaria parasites Any immune-interacting protein is potentially a vaccine target, but knowledge of global genetic diversity is an important requirement for vaccine development. For example, influenza and SARS-CoV-2 vaccines are developed based on the knowledge of how their genomes have evolved. However, the very unusual hotspots of genetic diversity in the P. falciparum DBLMSP and DBLMSP2 genes are so extreme that current algorithms for mapping genetic variants failed to capture them, leaving researchers unaware of a large proportion of the variation in these genes. To address this, the researchers developed new bioinformatics software that uses genome graphs and analysed a broad sample of parasites from 29 countries. This new approach revealed a wide range of previously hidden variants, and with these, they were able to demonstrate that multiple gene conversion events had occurred. These new variants, available for download from the website linked to the study, provide a valuable resource for the malaria research community. "Genome graphs are a great bioinformatics method to help us decode the complex genetic landscapes arising from the interplay between pathogens and human hosts," said Sorina Maciuca, co-author and former PhD student in the Iqbal group and Genomics Data Scientist at Genomics England. "They allow us to take into account a broader spectrum of genetic diversity and obtain new insights into how pathogens like P. falciparum evolve and evade our immune defences." What are genome graphs? The traditional approach in genomics is to define one reference genome and describe any other genome as a set of small differences from this reference. This does not work well when genomes differ too much. Genome graphs take a population of genomes and build an ensemble reference which is aware of all of the genetic variation in the species. "This research provides a comprehensive map of genetic diversity of these two fascinating genes in P. falciparum," said Zamin Iqbal, Group Leader at EMBL-EBI and Professor of Algorithmic and Microbial Genomics at the University of Bath. "We have been trying to understand the unusual patterns in these genes for almost a decade now, and our best hypothesis had been that the really different "versions'' of the gene were being preserved by natural selection, for unknown reasons. We have shown here that, in fact, this copying mechanism -- gene conversion -- has been repeatedly creating these anomalous different "versions" of the genes. This data not only enhances our grasp of malaria's biology, but also will be valuable to researchers across the world studying these genes and their interaction with our immune system."

A team of researchers is delving deeper into the science behind how spiny mice can regenerate lost tissue and using what they learn to trigger regeneration in other types of mice -- advances which one day may be translated into humans. Whereas adult laboratory mice heal injuries with scar tissue, spiny mice have the unique ability to regrow lost skin and regenerate musculoskeletal tissues in their body. A team of researchers at the University of Kentucky and Cincinnati Children's Hospital is delving deeper into the science behind how spiny mice can regenerate lost tissue and using what they learn to trigger regeneration in other types of mice -- advances which one day may be translated into humans. Whereas adult laboratory mice heal injuries with scar tissue, spiny mice have the unique ability to regrow lost skin and regenerate musculoskeletal tissues in their body. Ashley W. Seifert, Ph.D., an associate professor in the Department of Biology in the UK College of Arts and Sciences, and his research group have pioneered using spiny mice and other animal models to understand how complex tissue can regenerate, bridging regenerative biology and medicine. In April 2023, Science Advances published a study from Seifert's group focused on the cellular response to injury during tissue healing. The paper showed how ERK signaling acts as a key switch balancing the healing response, and it demonstrated that increasing and maintaining this type of signaling in laboratory mice could trigger a regenerative response instead of scarring. Their latest study, published in Developmental Cell, provides new insight into how specific immune cells respond to injury and direct tissue regeneration. The team of scientists looked at a specific type of immune cell known as macrophages, which play a crucial role in regulating the inflammatory response by defending injured tissue against pathogens while simultaneously promoting tissue repair in both spiny mice and lab mice. "In our study, we used a dual-species system to dissect macrophage phenotypes between two modes of healing: tissue regeneration or scar tissue formation," said Seifert. "While it's still unclear the extent to which macrophages ultimately control different healing trajectories, we observed subtle and distinct macrophage signatures that aligned specifically with regeneration." Using an identical injury in both kinds of rodents, the researchers studied signals released by macrophages that communicate to nearby cells, telling them to rebuild all the tissues that are missing. They also noted the characteristics, or phenotypes, of the cells during those processes. "Our study shows that spiny mouse macrophages release distinct proteins that are partially responsible for the reformation of specialized tissues at the site of injury and for protecting cells from stress," said Seifert. "Overall, our experiments point towards macrophages establishing a tissue microenvironment conducive for regeneration." Researchers examined and compared macrophages from the bone marrow of both types of mice. Then they used RNA sequencing to identify proteins these cells secrete. (RNA, or ribonucleic acid, is a molecule that carries genetic instructions from DNA to create proteins.) Researchers found that the macrophages from bone marrow responded differently on their own to interferon gamma, a protein produced by cells in response to infection, and lipopolysaccharide, a molecule found in some bacteria that stimulates the immune system. "In response to inflammatory cues, macrophages in spiny mice release unique communicating proteins that promote the growth of new blood vessels, lymphatic vessels, anti-inflammatory activities and tissue rebuilding that positively contribute to regenerative healing," said Seifert. The UK team found that those cells maintain their uniqueness throughout scar tissue formation or regeneration. Scientists studied samples at five different time points during healing to better understand the timeline for how macrophages do their work. A gene analysis showed researchers which types of molecules produced by the cells were most active in each type of mice and with which functions those were most associated. "In our work to answer more scientific questions about spiny mice, we found that a specific type of protein -- vascular endothelial growth factor c or VEGFC -- is uniquely secreted by spiny mouse macrophages during regeneration," said Seifert. "It also plays a multifunctional role, encouraging the growth of new blood and lymphatic vessels." "Through specific antibody blocking of VEGFC protein in the ears of spiny mice, we observed notable changes in the formation of blood vessels, lymph vessels and cell division. This also led to decreases in new hair follicle formation and increases in inflammation, ultimately disrupting the process of tissue regeneration," said Ajoy Aloysius, Ph.D., co-first author of the study and a postdoctoral scholar in biology. "Additionally, we speculate that enhancing the secretion of VEGFC and other factors by macrophages during fibrotic wound healing may contribute to a regenerative outcome in tissue healing, however further experiments are required to confirm this idea." The study ultimately suggests that macrophages found in specific tissues throughout the body help direct and regulate the cellular repair program. Importantly, the results of this study show that the type of macrophage matters and suggest that altering what one type of macrophage secretes could alter how tissue repairs itself. "This study aims to unlock the body's natural potential to regenerate after traumatic injury," said Jennifer Simkin, Ph.D., one of the study's first authors and an assistant professor of orthopaedic surgery at the Louisiana State University Health Sciences Center. "To do this, we study animals that can regenerate multiple tissues (hair, cartilage, muscle and skin) after injury to find the cellular and molecular signals necessary for better healing. Ultimately, we hope these findings pave the way for novel therapies to enhance wound healing." Researchers say further study is needed to better understand and define some of this cellular crosstalk during the regeneration process. This team of researchers represents the UK Departments of Biology and Physiology, the UK Spinal Cord and Brain Injury Research Center, Louisiana State University Health-New Orleans, the University of Cincinnati Children's Hospital Medical Center and the European Bioinformatics Institute.