Skip to main content
Most recent
This workflow lets you find potential circular DNA in your organism of interest using short-read, whole-genome sequencing data and a reference genome. We applied it to parasitoid wasps and some other parasites and found putative circular DNA.
We assembled a comprehensive E. coli antimicrobial resistance phenotypes-genotype resource. This dataset will aid large-scale genetic studies on anti-microbial resistance and support research in phylogenetics and other fields.
Inspired by wasps co-opting viral capsids to deliver genes to the caterpillars they parasitize, we looked for capsid-like proteins in other species. We found capsid homologs in ticks and other parasites, suggesting this phenomenon could be wider spread than previously known.
We explored the use of embeddings from protein language models to distinguish between genuine and putative coding open reading frames (ORFs). We found that an embeddings-based approach (shared as a small Python package called plm-utils) improves identification of short ORFs.
Data sets
We assembled a comprehensive E. coli antimicrobial resistance phenotypes-genotype resource. This dataset will aid large-scale genetic studies on anti-microbial resistance and support research in phylogenetics and other fields.
We previously released a draft genome assembly for the lone star tick, A. americanum. We've now predicted genes from this assembly to use for downstream functional characterization and comparative genomics efforts.
Seeing how microbes are organized within a community can inspire hypotheses about how species interact with each other. We used HiPR-FISH spatial imaging to look at the distribution of microbes within five distinct microbial communities growing on the surface of aged cheeses.
We generated a whole-genome assembly for the lone star tick to serve as a reference for downstream efforts where whole-genome maps are required. We created our assembly using pooled DNA from salivary glands of 50 adult female ticks that we sequenced using PacBio HiFi reads.
How do you approach getting a microbiome set up in a new lab? We’re sharing protocols for how we collected, stocked, and sequenced a set of cheese rind microbiomes and generated a high-quality metagenomics resource for future computational studies.
The way you generate a reference database has a real impact on the completeness and results of proteomics experiments.
Ideas
Oligo pools can contain millions of unique sequences, but they’re limited by length, error rate, and bias. We propose methods to scalably screen synthetic DNA libraries, so an individual researcher can obtain thousands of error-free synthetic DNA assemblies at low cost.
We investigated protein structure predictions as an alternative to protein sequence homology for comparing single-cell RNA-seq data across species.
The process of deciding whether a candidate actin homolog represents a “true” actin is tricky. We propose clear and data-driven criteria to define actin that highlight the functional importance of this protein while accounting for phylogenetic diversity.
Methods
This workflow lets you find potential circular DNA in your organism of interest using short-read, whole-genome sequencing data and a reference genome. We applied it to parasitoid wasps and some other parasites and found putative circular DNA.
This pub details a process for phage amplification and concentration, DNA extraction, and HPLC and MS analysis of phage nucleosides. We optimized the approach with model phages known to use non-canonical nucleosides in their DNA, but plan to apply it for other phages.
If you’re interested in generating proteomics data but your organism of interest doesn’t have a sequenced genome to use as a reference database, it is straightforward and useful to collect a transcriptome instead.
Negative data
We outline a comparative approach to investigate protein function by correlating the presence or absence of a protein with species-level phenotypes. We applied this strategy to a novel actin isoform in fungi but didn’t find an association with any of the phenotypes we considered.
We struggled to isolate enough phage mRNA for HPLC as we searched for new nucleoside chemistries. Ribodepleting the abundant extracted rRNA introduced contaminating DNA, and we were still left with more bacterial mRNA than phage transcripts. We suggest an alternative approach.
Open questions
We’re using the well-studied superfamily of small monomeric GTPases, the Ras GTPases, to evaluate our structure-based clustering tool, ProteinCartography. We’re seeking feedback on working with this protein family and determining which individual proteins to study.
How can we biochemically validate protein function predictions with the deoxycytidine kinase family?
The human deoxycytidine kinase, a member of the nucleoside salvage pathway, has been studied extensively. We’ll use this family to assess our structure-based protein clustering tool, ProteinCartography. We’d love feedback on how we might work with this protein for validation.
We aim to validate ProteinCartography, a tool for structure-based protein clustering, by evaluating two foundational hypotheses: that proteins within a cluster have similar functions and proteins in different clusters have differing functions.
How can we measure the true impact of science? We're seeking feedback on indicators of the utility and rigor of publications beyond traditional journal metrics. Your input will help shape the future of our publishing experiment.
We’re seeking feedback on NovelTree, our modular phylogenomic workflow. We’d appreciate your insights into how we can improve gene family inference, incorporate protein structure predictions, and expand to whole-genome data as input.
Since releasing our pub on polyphosphate-forming proteins in bacteria, we’ve noticed the community has similar problems studying this process in diverse organisms. We’re actively seeking feedback with a focus on advancing basic discoveries and useful tools in this space!
Perspectives
Two years into our publishing experiment, we’ve learned a lot. We built internal processes that worked but inadvertently decreased scientists' agency and creativity. Now, we’re minimizing process in an effort to empower our scientists to share their work how they see fit.
Our "icebox" is where we share the projects that we've decided not to continue. We surveyed our company to understand resulting cultural shifts and how others are carrying our iced work forward.
Genetic models of complex traits often rely on incorrect assumptions that drivers of trait variation are additive and independent. An information theoretic framework for analyzing trait variation can better capture phenomena like allelic dominance and gene-gene interaction.
A workflow orchestration framework can streamline repeatable tasks and make workflows broadly usable. From several options, we chose Nextflow due to the ease of deploying across platforms, vibrant nf-core community, and ability to manage and monitor workflows with Nextflow Tower.
Building on the open-source platform PubPub, we’re sharing the first iteration of our publishing website. In addition to posting our first set of research pubs, we’re documenting our progress in developing this new system for sharing science and hope you’ll provide feedback.
Resources
We explored the use of embeddings from protein language models to distinguish between genuine and putative coding open reading frames (ORFs). We found that an embeddings-based approach (shared as a small Python package called plm-utils) improves identification of short ORFs.
Peptigate predicts bioactive peptides from transcriptomes. It integrates existing tools to predict sORF-encoded peptides, cleavage peptides, and RiPPs, then annotates them for bioactivity and other properties. We welcome feedback on expanding its capabilities.
We’ve developed an easy-to-assemble apparatus and software for the automated collection of visible biological phenotypes such as growth, macroscopic morphology, motion, reflectance, and fluorescence.
Machine learning is a powerful tool for classifying images in a time series, such as the developmental stages of embryos. We built a classifier using only bright-field microscopy images to infer nematode embryonic stages at high throughput.
We distilled label-free microscopy data by comparing and implementing feature-detection algorithms. Sobel and Laplacian methods outperformed pixel intensity variance in accuracy.
The ProteinCartography pipeline identifies proteins related to a query protein using sequence- and structure-based searches, compares all protein structures, and creates a navigable map that can be used to look at protein relationships and make hypotheses about function.
We want to find and use evolutionary innovations to solve present-day problems. We developed NovelTree, an efficient phylogenomic workflow that will empower us to decode the evolutionary traces of these innovations across the tree of life.
Horizontal gene transfer (HGT) is the exchange of DNA between species. It can lead to the acquisition of new gene functions, so finding HGT events can reveal genome novelty. preHGT is a pipeline that uses multiple existing methods to quickly screen for transferred genes.
The increasingly large number of sequences available in public databases makes searches slower and slower. We clustered the NCBI non-redundant protein database and calculated taxonomic info for each cluster. This collapses similar sequences and reduces the database by over half.
We want to seamlessly process and summarize metagenomics data from Illumina or Nanopore technologies. We built a Nextflow workflow that handles common metagenomics tasks and produces useful outputs and intuitive visualizations.
Constraining motile microorganisms for live imaging often requires costly microfluidics or optical traps to keep them in view. We used patterned stamps and agar to make versatile, inexpensive “microchambers” and offer a way to predict the right chamber size for a given organism.
We want to swiftly generate genome assemblies and produce quality control statistics to gauge the need for more curation. We built a Nextflow pipeline that assembles Illumina, Nanopore, or PacBio sequencing reads for a single organism and runs QC checks on the resulting assembly.
seqqc is a Nextflow pipeline for quality control of short- or long-read sequencing data. It quickly assesses the quality of sequencing data so that it can be posted to a public repository before analysis for biological insights. Faster open data, faster knowledge for everyone.
The sourmash Python package produces many outputs that describe the content and similarity of sequencing data. We developed a new R package, sourmashconsumr, that lets a wider range of users easily load, analyze, and visualize those outputs in R.
Cells can be highly motile, moving in and out of a microscope’s field of view. Understanding complex life cycles is difficult without continuous observation. To overcome this challenge, we’ve developed a 3D-printed microchamber device to confine cells for long-term visualization.
Researchers studying any organism with genomic data can follow this simple walkthrough to create sets of barcoded probes for the multiplexed FISH technique called MERFISH. We’re sharing interactive code notebooks that can be adapted to design barcoded FISH probes for any species.
Results
Many protein prediction and design models rely on evolutionary comparisons. We show that popular databases are phylogenetically biased, influencing the statistical utility of the known protein universe in important ways.
Sensory disorders are clinically common, debilitating conditions. But mouse behavioral models are often insufficient. We demonstrate that label-free, minimally-invasive brain imaging in mice could be a promising avenue for sensory research or drug discovery efforts.
To test its utility in analyzing biological samples, we built an open-source Raman spectrometer and collected spectra from chilis, beer, and algae. We could stratify samples, classify replicates, and link spectra with quantitative traits of beer (ABV) and chilis (perceived heat).
Inspired by wasps co-opting viral capsids to deliver genes to the caterpillars they parasitize, we looked for capsid-like proteins in other species. We found capsid homologs in ticks and other parasites, suggesting this phenomenon could be wider spread than previously known.
We've applied ProteinCartography, a tool for protein family exploration, to the well-studied actin family. We’re able to categorize actins and related proteins into distinguishable functional buckets, and we uncovered some surprising hypotheses that could prompt further study.
Only some bacteria accumulate substantial amounts of polyphosphate (polyP). We thought that despite sequence divergence, polyP synthesis enzymes in these bacteria might have similar structures. We found this is sometimes true but doesn’t fully explain the phenomenon.
It is commonly assumed that phenotypes arise from the cumulative effects of many independent genes. However, we show that by accounting for dependent and nonlinear biological relationships, we can generate models that predict phenotypes with great accuracy.
We implemented a lightweight method to identify viruses in 342 human brain bulk and single-cell sequencing data sets, and identified two glioblastoma cells from a single patient that contained deltapolyomavirus sequences.
Some human proteins are encoded by genes with repetitive sequences, which, if they expand, damage the nervous system and cause disorders like Huntington’s disease. We found animals with similar proteins that have more repeats than we’ve ever seen in healthy people.
We sampled cheese microbial communities to discover bacteriophages with unusual genome chemistries. We isolated 114 bacterial host strains and 17 phages, and identified one phage with a probable arabinose hypermodification of hydroxymethylcytosine.
We are interested in neuroactive metabolites that influence animal behavior. Some fungi have horizontally transferred neuroactive metabolite pathways between species. We used a horizontal gene transfer detection pipeline to screen for novel fungal genes tied to neuroactivity.
We’re crossing C. reinhardtii and C. smithii algae for high-throughput genotype-phenotype mapping. In preparation, we’re comparing the parents to uncover unique species-specific phenotypes.
Hoping to find proteins that alter physiology in useful ways, we screened venom data sets for toxins fused to domains with additional functionality. We identified candidates, but struggled to infer any novel functions, and none seem well-conserved across venomous species.
Even with many tools available, categorizing species is tough. We used data from Raman spectroscopy, a form of label-free imaging, to infer phylogenetic patterns among several dozen diverse microbial taxa, offering a non-destructive and rapid way to dissect species relationships.
Long protrusions from several microalgal species appear to help cells move, capture prey, transport mitochondria and chloroplasts, and more. Are they filopodia that evolved abilities more like other actin- or microtubule-based structures, or are they something new?
Quantifying movement is a powerful window into cellular functions. However, cells can generate movement through a variety of complex mechanisms. Here, we generate a flexible framework for comparing an especially variable type of motility: cellular crawling.
Treating P. tricornutum cells with serine endopeptidases or certain cytoskeletal inhibitors induces the formation of cell wall-free protoplasts and suggests a novel role for actin and myosin in preventing protoplast formation.
All pubs
Many protein prediction and design models rely on evolutionary comparisons. We show that popular databases are phylogenetically biased, influencing the statistical utility of the known protein universe in important ways.
This workflow lets you find potential circular DNA in your organism of interest using short-read, whole-genome sequencing data and a reference genome. We applied it to parasitoid wasps and some other parasites and found putative circular DNA.
We assembled a comprehensive E. coli antimicrobial resistance phenotypes-genotype resource. This dataset will aid large-scale genetic studies on anti-microbial resistance and support research in phylogenetics and other fields.
We explored the use of embeddings from protein language models to distinguish between genuine and putative coding open reading frames (ORFs). We found that an embeddings-based approach (shared as a small Python package called plm-utils) improves identification of short ORFs.
Sensory disorders are clinically common, debilitating conditions. But mouse behavioral models are often insufficient. We demonstrate that label-free, minimally-invasive brain imaging in mice could be a promising avenue for sensory research or drug discovery efforts.
Two years into our publishing experiment, we’ve learned a lot. We built internal processes that worked but inadvertently decreased scientists' agency and creativity. Now, we’re minimizing process in an effort to empower our scientists to share their work how they see fit.
Peptigate predicts bioactive peptides from transcriptomes. It integrates existing tools to predict sORF-encoded peptides, cleavage peptides, and RiPPs, then annotates them for bioactivity and other properties. We welcome feedback on expanding its capabilities.
We’re using the well-studied superfamily of small monomeric GTPases, the Ras GTPases, to evaluate our structure-based clustering tool, ProteinCartography. We’re seeking feedback on working with this protein family and determining which individual proteins to study.
How can we biochemically validate protein function predictions with the deoxycytidine kinase family?
The human deoxycytidine kinase, a member of the nucleoside salvage pathway, has been studied extensively. We’ll use this family to assess our structure-based protein clustering tool, ProteinCartography. We’d love feedback on how we might work with this protein for validation.
We aim to validate ProteinCartography, a tool for structure-based protein clustering, by evaluating two foundational hypotheses: that proteins within a cluster have similar functions and proteins in different clusters have differing functions.
To test its utility in analyzing biological samples, we built an open-source Raman spectrometer and collected spectra from chilis, beer, and algae. We could stratify samples, classify replicates, and link spectra with quantitative traits of beer (ABV) and chilis (perceived heat).
How can we measure the true impact of science? We're seeking feedback on indicators of the utility and rigor of publications beyond traditional journal metrics. Your input will help shape the future of our publishing experiment.
Inspired by wasps co-opting viral capsids to deliver genes to the caterpillars they parasitize, we looked for capsid-like proteins in other species. We found capsid homologs in ticks and other parasites, suggesting this phenomenon could be wider spread than previously known.
We’re seeking feedback on NovelTree, our modular phylogenomic workflow. We’d appreciate your insights into how we can improve gene family inference, incorporate protein structure predictions, and expand to whole-genome data as input.
We’ve developed an easy-to-assemble apparatus and software for the automated collection of visible biological phenotypes such as growth, macroscopic morphology, motion, reflectance, and fluorescence.
Oligo pools can contain millions of unique sequences, but they’re limited by length, error rate, and bias. We propose methods to scalably screen synthetic DNA libraries, so an individual researcher can obtain thousands of error-free synthetic DNA assemblies at low cost.
Since releasing our pub on polyphosphate-forming proteins in bacteria, we’ve noticed the community has similar problems studying this process in diverse organisms. We’re actively seeking feedback with a focus on advancing basic discoveries and useful tools in this space!
We outline a comparative approach to investigate protein function by correlating the presence or absence of a protein with species-level phenotypes. We applied this strategy to a novel actin isoform in fungi but didn’t find an association with any of the phenotypes we considered.
We've applied ProteinCartography, a tool for protein family exploration, to the well-studied actin family. We’re able to categorize actins and related proteins into distinguishable functional buckets, and we uncovered some surprising hypotheses that could prompt further study.
We previously released a draft genome assembly for the lone star tick, A. americanum. We've now predicted genes from this assembly to use for downstream functional characterization and comparative genomics efforts.
Machine learning is a powerful tool for classifying images in a time series, such as the developmental stages of embryos. We built a classifier using only bright-field microscopy images to infer nematode embryonic stages at high throughput.
Our "icebox" is where we share the projects that we've decided not to continue. We surveyed our company to understand resulting cultural shifts and how others are carrying our iced work forward.
We distilled label-free microscopy data by comparing and implementing feature-detection algorithms. Sobel and Laplacian methods outperformed pixel intensity variance in accuracy.
The ProteinCartography pipeline identifies proteins related to a query protein using sequence- and structure-based searches, compares all protein structures, and creates a navigable map that can be used to look at protein relationships and make hypotheses about function.
Only some bacteria accumulate substantial amounts of polyphosphate (polyP). We thought that despite sequence divergence, polyP synthesis enzymes in these bacteria might have similar structures. We found this is sometimes true but doesn’t fully explain the phenomenon.
Genetic models of complex traits often rely on incorrect assumptions that drivers of trait variation are additive and independent. An information theoretic framework for analyzing trait variation can better capture phenomena like allelic dominance and gene-gene interaction.
It is commonly assumed that phenotypes arise from the cumulative effects of many independent genes. However, we show that by accounting for dependent and nonlinear biological relationships, we can generate models that predict phenotypes with great accuracy.
We implemented a lightweight method to identify viruses in 342 human brain bulk and single-cell sequencing data sets, and identified two glioblastoma cells from a single patient that contained deltapolyomavirus sequences.
Some human proteins are encoded by genes with repetitive sequences, which, if they expand, damage the nervous system and cause disorders like Huntington’s disease. We found animals with similar proteins that have more repeats than we’ve ever seen in healthy people.
We want to find and use evolutionary innovations to solve present-day problems. We developed NovelTree, an efficient phylogenomic workflow that will empower us to decode the evolutionary traces of these innovations across the tree of life.
We sampled cheese microbial communities to discover bacteriophages with unusual genome chemistries. We isolated 114 bacterial host strains and 17 phages, and identified one phage with a probable arabinose hypermodification of hydroxymethylcytosine.
We are interested in neuroactive metabolites that influence animal behavior. Some fungi have horizontally transferred neuroactive metabolite pathways between species. We used a horizontal gene transfer detection pipeline to screen for novel fungal genes tied to neuroactivity.
Horizontal gene transfer (HGT) is the exchange of DNA between species. It can lead to the acquisition of new gene functions, so finding HGT events can reveal genome novelty. preHGT is a pipeline that uses multiple existing methods to quickly screen for transferred genes.
We investigated protein structure predictions as an alternative to protein sequence homology for comparing single-cell RNA-seq data across species.
The increasingly large number of sequences available in public databases makes searches slower and slower. We clustered the NCBI non-redundant protein database and calculated taxonomic info for each cluster. This collapses similar sequences and reduces the database by over half.
We’re crossing C. reinhardtii and C. smithii algae for high-throughput genotype-phenotype mapping. In preparation, we’re comparing the parents to uncover unique species-specific phenotypes.
Hoping to find proteins that alter physiology in useful ways, we screened venom data sets for toxins fused to domains with additional functionality. We identified candidates, but struggled to infer any novel functions, and none seem well-conserved across venomous species.
We want to seamlessly process and summarize metagenomics data from Illumina or Nanopore technologies. We built a Nextflow workflow that handles common metagenomics tasks and produces useful outputs and intuitive visualizations.
Constraining motile microorganisms for live imaging often requires costly microfluidics or optical traps to keep them in view. We used patterned stamps and agar to make versatile, inexpensive “microchambers” and offer a way to predict the right chamber size for a given organism.
Even with many tools available, categorizing species is tough. We used data from Raman spectroscopy, a form of label-free imaging, to infer phylogenetic patterns among several dozen diverse microbial taxa, offering a non-destructive and rapid way to dissect species relationships.
Seeing how microbes are organized within a community can inspire hypotheses about how species interact with each other. We used HiPR-FISH spatial imaging to look at the distribution of microbes within five distinct microbial communities growing on the surface of aged cheeses.
We want to swiftly generate genome assemblies and produce quality control statistics to gauge the need for more curation. We built a Nextflow pipeline that assembles Illumina, Nanopore, or PacBio sequencing reads for a single organism and runs QC checks on the resulting assembly.
We generated a whole-genome assembly for the lone star tick to serve as a reference for downstream efforts where whole-genome maps are required. We created our assembly using pooled DNA from salivary glands of 50 adult female ticks that we sequenced using PacBio HiFi reads.
We struggled to isolate enough phage mRNA for HPLC as we searched for new nucleoside chemistries. Ribodepleting the abundant extracted rRNA introduced contaminating DNA, and we were still left with more bacterial mRNA than phage transcripts. We suggest an alternative approach.
How do you approach getting a microbiome set up in a new lab? We’re sharing protocols for how we collected, stocked, and sequenced a set of cheese rind microbiomes and generated a high-quality metagenomics resource for future computational studies.
A workflow orchestration framework can streamline repeatable tasks and make workflows broadly usable. From several options, we chose Nextflow due to the ease of deploying across platforms, vibrant nf-core community, and ability to manage and monitor workflows with Nextflow Tower.
seqqc is a Nextflow pipeline for quality control of short- or long-read sequencing data. It quickly assesses the quality of sequencing data so that it can be posted to a public repository before analysis for biological insights. Faster open data, faster knowledge for everyone.
The sourmash Python package produces many outputs that describe the content and similarity of sequencing data. We developed a new R package, sourmashconsumr, that lets a wider range of users easily load, analyze, and visualize those outputs in R.
This pub details a process for phage amplification and concentration, DNA extraction, and HPLC and MS analysis of phage nucleosides. We optimized the approach with model phages known to use non-canonical nucleosides in their DNA, but plan to apply it for other phages.
Long protrusions from several microalgal species appear to help cells move, capture prey, transport mitochondria and chloroplasts, and more. Are they filopodia that evolved abilities more like other actin- or microtubule-based structures, or are they something new?
The process of deciding whether a candidate actin homolog represents a “true” actin is tricky. We propose clear and data-driven criteria to define actin that highlight the functional importance of this protein while accounting for phylogenetic diversity.
Quantifying movement is a powerful window into cellular functions. However, cells can generate movement through a variety of complex mechanisms. Here, we generate a flexible framework for comparing an especially variable type of motility: cellular crawling.
Cells can be highly motile, moving in and out of a microscope’s field of view. Understanding complex life cycles is difficult without continuous observation. To overcome this challenge, we’ve developed a 3D-printed microchamber device to confine cells for long-term visualization.
Building on the open-source platform PubPub, we’re sharing the first iteration of our publishing website. In addition to posting our first set of research pubs, we’re documenting our progress in developing this new system for sharing science and hope you’ll provide feedback.
Researchers studying any organism with genomic data can follow this simple walkthrough to create sets of barcoded probes for the multiplexed FISH technique called MERFISH. We’re sharing interactive code notebooks that can be adapted to design barcoded FISH probes for any species.
Treating P. tricornutum cells with serine endopeptidases or certain cytoskeletal inhibitors induces the formation of cell wall-free protoplasts and suggests a novel role for actin and myosin in preventing protoplast formation.
The way you generate a reference database has a real impact on the completeness and results of proteomics experiments.
If you’re interested in generating proteomics data but your organism of interest doesn’t have a sequenced genome to use as a reference database, it is straightforward and useful to collect a transcriptome instead.
