Unveiling the Hidden RNA Virosphere: A Deep Dive into Novel Virus Discovery
0:00:03 Introduction to fundraising efforts and appeal for listener support to continue producing science-focused programs.
0:01:03 This is episode 1179 of the podcast This Week in Virology (TWiV). The hosts introduce themselves and provide weather updates.
0:04:24 Discussion on the importance of listener support for TWiV, a non-profit, ad-free podcast.
0:05:25 Update on a previous video regarding RFK Jr.'s unsuitability for a government role, with a corrected link to a website for contacting senators.
0:06:58 Job opportunity at the Rosenfeld Laboratory within the FDA's Center for Biologics Evaluation and Research (CBER) for a postdoctoral staff scientist to study immune responses during adenovirus infection.
0:07:24 Announcement of the American Society for Virology (ASV) 2025 meeting in Montreal, Canada, with details on registration, membership renewal, abstract submission, and travel awards.
0:09:28 Discussion about potentially hosting a TWiV episode in French at the ASV 2025 meeting.
0:10:58 News item: RFK Jr.'s lawyer petitions the FDA to revoke the polio vaccine license, despite its proven safety and efficacy.
0:13:48 News item: Sequencing data from the Wuhan Institute of Virology shows no close relatives to SARS-CoV-2, supporting the natural origin hypothesis.
0:16:15 Snippet: Introduction to a paper on using artificial intelligence to discover novel RNA viruses.
0:16:49 The "dark virome" or "biological dark matter" refers to viruses that exist but are difficult to characterize using traditional methods like PCR.
0:17:28 A new machine learning algorithm (AI) was developed that analyzes both RNA sequence data and predicted protein structures of RNA-directed RNA polymerases (essential for RNA virus replication).
0:17:50 The AI was trained on metagenomic data (nucleotide sequence information from environmental samples) and identified numerous novel RNA virus genomes.
0:18:37 The AI is unique because it incorporates structural predictions of viral proteins, allowing it to identify highly divergent viruses that would be missed by sequence-based methods alone.
0:19:04 The researchers validated their AI predictions by pulling out and sequencing some of the novel viral genomes, confirming their existence.
0:19:24 The discovered viruses come from diverse environments, including extreme ones. This AI approach can help illuminate the "dark virosphere" and provide a trove of new viral genomes for further study.
0:19:51 Summary: Researchers used AI to analyze transcriptomes (not genomes) and identified many new viruses.
0:20:11 The paper's authors and their affiliations are listed, highlighting the collaborative nature of the research.
0:21:09 RNA viruses are commonly found in various environments and often studied by looking for RNA-dependent RNA polymerase.
0:21:32 The paper uses AI, specifically deep learning algorithms and Transformer architecture, to identify potential RNA viruses in a large dataset of metatranscriptomic data.
0:23:44 Two strategies were used: an AI algorithm called LucaPro and a more conventional sequence-based approach.
0:24:49 LucaPro identified over 160,000 putative viral species, including 60 previously unidentified groups.
0:25:39 The researchers performed various analyses to ensure the accuracy of LucaPro and confirm that the identified sequences were indeed from RNA viruses.
0:27:31 Some of the newly discovered RNA viruses have very long genomes and contain other proteins with unknown functions.
0:28:25 The AI approach used in this study is agnostic, meaning it doesn't rely on prior knowledge of viral hosts or specific viral proteins.
0:29:03 The study significantly expands the known diversity of RNA viruses, increasing the number of known species by 55.9-fold.
0:29:38 The greatest diversity of novel viruses was found in aquatic environments, particularly in leaf litter, wetlands, freshwater, and wastewater.
0:31:23 The Antarctic samples, collected in January and February 2022, also yielded a significant number of novel viruses.
0:31:45 The researchers suggest that the focus on viruses that cause human disease might have led to a bias in previous sampling efforts.
0:32:39 The AI framework developed in this study could be adapted to discover other types of biological dark matter.
0:33:13 The uniqueness of this approach lies in the incorporation of structural data, which allows for the identification of viruses with highly divergent sequences.
0:34:47 Discussion on the limitations of current AI models and the potential legal issues surrounding the use of copyrighted data for training AI.
0:35:55 Introduction to a new Google AI tool (NotebookLM) that can summarize scientific papers into a podcast format.
0:37:54 Discussion on the potential use of AI for summarizing scientific papers and the limitations of current AI models in understanding and interpreting complex scientific information.
0:39:54 Introduction to a paper on using biomarkers to discriminate between early and late phases of respiratory viral infections.
0:40:11 The interferon response, specifically the expression of interferon-stimulated genes (ISGs), can be used to detect viral infections early.
0:41:21 Previous research identified the ISG called IFI27 as a reliable indicator of SARS-CoV-2 infection.
0:41:54 This study used a controlled human challenge study with SARS-CoV-2 to precisely track the timing of ISG expression.
0:42:24 Ethical considerations of human challenge studies are briefly discussed.
0:42:35 The study found that IFI27 is turned on reliably during infection, and another ISG called MX1 is turned on even earlier, often before symptoms or PCR positivity.
0:43:06 The same pattern of MX1 and IFI27 expression was observed in previous challenge studies with influenza, RSV, and rhinovirus.
0:43:16 MX1 and IFI27 expression in blood cells could potentially be used as a diagnostic test for respiratory virus infections, even in asymptomatic cases.
0:44:05 Details of the SARS-CoV-2 challenge study, including participant demographics, virus inoculation method, and sample collection.
0:45:33 Discussion about the variability in individual responses to infection and the potential implications for understanding disease susceptibility.
0:46:58 Ethical concerns about the challenge study are raised, particularly regarding the potential risks to young, healthy volunteers.
0:48:47 Discussion on the potential use of MX1 and IFI27 levels to determine the timing of infection.
0:49:54 MX1 is found to be a better predictor of infectiousness than IFI27.
0:50:20 Single-cell RNA sequencing revealed that MX1 is expressed in various immune cell types, while IFI27 is restricted to myeloid cells.
0:50:48 Differences in chromatin openness around the MX1 and IFI27 genes might explain the faster induction of MX1.
0:53:31 The study suggests that combining MX1 and IFI27 measurements could provide a more comprehensive picture of the infection timeline.
0:55:35 Discussion on the limitations of using nasal swabs for detecting these biomarkers compared to blood samples.
0:56:47 Discussion on the potential use of these biomarkers to estimate the time since infection in naturally infected individuals.
0:57:30 MX1 levels might be elevated even in the absence of detectable virus replication, suggesting a rapid initial immune response.
0:58:54 Discussion on the potential use of these biomarkers to identify individuals who are infected but asymptomatic.
1:01:03 The authors propose that the key conceptual advance of this work is the link between epigenetic regulation and the rapid induction of MX1.
1:04:45 Discussion on the need for a "near-patient platform" (point-of-care test) to further evaluate the clinical utility of these biomarkers.
1:05:55 Discussion about the potential use of these biomarkers to differentiate between different types of infections (bacterial vs. viral) and to monitor response to vaccination.
1:07:54 Listener email about the potential for leaf blowers to spread infectious microbes.
1:09:11 Discussion about the potential risks of using hand dryers in public restrooms.
1:11:19 Listener email requesting help in locating a specific TWiV episode that discussed the relationship between science and politics.
1:13:15 The hosts identify the episode as TWiV 1155 and discuss the content of the listener letter.
1:14:18 Listener email providing information about New Zealand and suggesting a researcher to contact regarding bird flu.
1:15:14 Update on a reported outbreak of a respiratory virus in the DRC, which turned out to be malaria.
1:15:45 Listener email about the bumbling biochemist and a link to her blog.
1:16:07 Listener email praising TWiV and sharing a link to a German techno marching band called MEUTE.
1:19:17 Listener email about a DNA test revealing a surprising family secret.
1:21:54 Hosts' picks of the week: Angela picks an article about a new HIV drug, Kathy picks a video of a piano performance and an article about the return of the FAO Schwarz big piano, Rich picks the book and movie Never Cry Wolf, and Vincent picks a webpage about the differences between ravens and crows and a NASA rocket engine fireplace video loop.