Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences

Zhu, Anqi; Ibrahim, Joseph G.; Love, Michael I.

doi:10.1101/303255

Cited by 318 publications

(343 citation statements)

References 36 publications

Supporting

Mentioning

342

Contrasting

Unclassified

Order By: Relevance

“…Genes with less than ten reads in more than 75% of the samples were removed. Genes were tested for differential expression using DESeq2 v1.18.1 (Love et al, 2014) with shrinkage estimator apeglm v1.0.3 (Zhu et al, 2018) (Table S6). Gene Ontology category GO:0006955 was used to classify immune response genes.…”

Section: Star Methodsmentioning

confidence: 99%

Comparative Flavivirus-Host Protein Interaction Mapping Reveals Mechanisms of Dengue and Zika Virus Pathogenesis

Shah

Link

Jang

et al. 2018

Cell

270

355

View full text Add to dashboard Cite

Mosquito-borne flaviviruses, including dengue virus (DENV) and Zika virus (ZIKV), are a growing public health concern. Systems level analysis of how flaviviruses hijack cellular processes through virus-host protein-protein interactions (PPIs) provide information about their replication and pathogenic mechanisms. We used affinity purification-mass spectrometry (AP-MS) to compare flavivirus-host interactions for two viruses (DENV and ZIKV) in two hosts (human and mosquito). Conserved virus-host PPIs revealed that the flavivirus NS5 protein suppresses interferon stimulated genes by inhibiting recruitment of the transcription complex PAF1C, and that chemical modulation of SEC61 inhibits DENV and ZIKV replication in human and mosquito cells. Finally, we identified a ZIKV-specific interaction between NS4A and ANKLE2, a gene linked to hereditary microcephaly, and showed that ZIKV NS4A causes microcephaly in Drosophila in an ANKLE2-dependent manner. Thus, comparative flavivirus-host PPI mapping provides biological insights, and when coupled with in vivo models, can be used to unravel pathogenic mechanisms.

show abstract

Section: Star Methodsmentioning

confidence: 99%

Comparative Flavivirus-Host Protein Interaction Mapping Reveals Mechanisms of Dengue and Zika Virus Pathogenesis

Shah

Link

Jang

et al. 2018

Cell

270

355

View full text Add to dashboard Cite

show abstract

“…Specifically, differential expression was performed for each pair of flasks describing the before and after of an ALE experiment. We utilized an adaptive t prior shrinkage estimator (Zhu, Ibrahim and Love, 2018) to transform the log fold changes for better ranking and visualization of the differential expression results. Scatter plots of differential expression levels utilized the shrinked log fold changes ( Supplementary Fig.…”

Section: Differential Expression Analysis Of Rna-seqmentioning

confidence: 99%

Laboratory evolution of multipleE. colistrains reveals unifying principles of adaptation but diversity in driving genotypes

Kavvas

Feist

Antoniewicz

et al. 2020

Preprint

View full text Add to dashboard Cite

Fitness landscapes are a central concept in evolutionary biology and have been thoroughly detailed in terms of genotypes. However, our understanding of the selected metabolic and gene expression adaptations, and their dependence on genetic background, remains limited. Here, we reveal multi-scale adaptation principles in the E. coli species by taking multi-omics measurements of six different strains throughout their adaptive evolution to glucose minimal media. Statistics and matrix factorization is applied to yield four key results. First, analysis of the metabolic and physiological data shows evolutionary convergence in growth rate, glucose uptake rate, glycolytic ATP and NADH production but divergence in NADPH production strategies. Second, factorization-based analysis of the transcriptome revealed six conserved transcriptomic adaptations describing increased expression of ribosome and amino acid biosynthetic genes and decreased expression of stress response and structural genes. Third, correlation analysis identifies five tradeoffs underlying the transcriptomic profiles. Fourth, statistical tests leveraging ALE design identify four mutation-flux correlates and eight mutation-transcriptomic correlates that link mutations to systems level adaptation principles. Our total results reveal the dominant metabolic and regulatory constraints governing E. coli growth adaptation that either distinguish strains or are conserved principles.

show abstract

“…Filtering on all mapped gene counts was performed to exclude genes where the sum of counts in all the conditions was inferior to 10 counts. Default parameters were used with DESeq2 including the shrinks log2 fold-change (FC) estimated for each tested comparison [35,36]. A log 2 Fold Change and its standard error were generated in addition to a P-value (p-value) and a P-adj (Adjusted p-value) to account for the false discovery rate.…”

Section: Bioinformatic Pipeline and De Novo Assemblymentioning

confidence: 99%

Leishmania infection induces a limited differential gene expression in the sand fly midgut

Coutinho-Abreu

Serafim

Meneses

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: Phlebotomine sand flies are the vectors of Leishmania worldwide. To develop in the sand fly midgut, Leishmania multiplies and undergoes multiple stage differentiations leading to the infective form, the metacyclic promastigotes. To gain a better understanding of the influence of Leishmania infection on midgut gene expression, we performed RNA-Seq comparing uninfected and Leishmania infantum -infected Lutzomyia longipalpis midguts at seven time points which cover the various Leishmania developmental stages including early time points when blood digestion is taking place, and late time points when the parasites are undergoing metacyclogenesis. Results: Out of over 13,841 transcripts assembled de novo , only 113 sand fly transcripts, about 1%, were differentially expressed. Further, we observed a low overlap of differentially expressed sand fly transcripts across different time points suggesting that each Leishmania stage influences midgut gene expression in a specific manner. Two main patterns of sand fly gene expression modulation were noted. At early time points (days 1-4), more transcripts were down-regulated by Leishmania infection at large fold changes (> -32 fold). Among the down-regulated genes, the transcription factor Forkhead/HNF-3 and hormone degradation enzymes were differentially regulated on day 4 and appear to be the upstream regulators of nutrient transport, digestive enzymes, and peritrophic matrix proteins. Conversely, at later time points (days 6 onwards), most of the differentially expressed transcripts were up-regulated by small fold changes (< 32 fold). The molecular function of these genes has been associated with the metabolism of lipids and detoxification of xenobiotics. Conclusion: Overall, it appears that Leishmania modulates sand fly gene expression early on in order to overcome the barriers imposed by the midgut, yet it behaves like a commensal at later time points, where a massive number of parasites in the anterior midgut results only in modest changes in midgut gene expression.

show abstract

Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences

Cited by 318 publications

References 36 publications

Comparative Flavivirus-Host Protein Interaction Mapping Reveals Mechanisms of Dengue and Zika Virus Pathogenesis

Comparative Flavivirus-Host Protein Interaction Mapping Reveals Mechanisms of Dengue and Zika Virus Pathogenesis

Laboratory evolution of multipleE. colistrains reveals unifying principles of adaptation but diversity in driving genotypes

Leishmania infection induces a limited differential gene expression in the sand fly midgut

Contact Info

Product

Resources

About