Now that we’ve completed a lot of the foundational work for the E5 deep-dive expression project (e.g. sRNA & lncRNA ID, RNA and sRNA expression summaries, preliminary target prediction and coexpression work, etc.), I want to clarify our focus for project completion.
deep-dive-expression repo
manuscript
project meeting notes
hypotheses brainstorming document
Hypotheses:
Gene expression and sRNA
sRNA that bind to genes and are inversely correlated with expression - regulate those genes
There will be a suite of genes that a sRNA regulates that have similar function
Some lncRNAs that act as sponges (bind) sRNA will impact gene expression sRNA regulate
A subset of lncRNAs will regulate mRNA as evident by correlation
some lncRNAs may act as miRNA precursors – expect they will contain the full pre-miRNA
Lack of DNA methylation in a given will result in increased alternative isoforms
Gene expression and DNA methylation
Housekeeping genes: positive relationship between expression and GBM
plastic/environmentally responsive genes: negative relationship between expression and GBM
sRNA and DNA methylation
miRNAs modulate DNA methylation by binding to the mRNA encoding methylation machinery (e.g., DNMTs)
- For miRNA that target methylation machinery, expect negative relationship miRNA expression and global methylation
siRNA modulate DNA methylation through RNA-directed DNA methylation by binding to genomic region and recruiting protein machinery to direct methylation
- Expect positive relationship between siRNA binding in a genomic region and methylation of that region
siRNAs could also play a role in guiding DNA methylation to silence transposable elements
- Expect siRNA binding to TEs
Gene expression and histones
Data:
RNA sequencing
gene expression
lncRNA ID & expression
sRNA sequencing
miRNA ID & expression
siRNA ID & expression
piRNA ID & expression (if desired)
WGBS (in progress)
- DNA methylation mapping
Some H4 data
Completed work:
The below summary is for A.pulchra only.
All data processing for RNAseq and sRNAseq (including filtering, trimming, lncRNA ID, sRNA ID, summarizing to count matrices, and functional annotation) has been completed.
genes-ncRNA:
Two methods of miRNA-mRNA target prediction (miRanda and RNAhybrid) have been performed, and the miRanda results have been used in combination with Pearson’s Correlation to generate coexpression/interaction network. This interaction network has also been functionally annotated.
genes-WGBS:
none
ncRNA-WGBS:
I’ve looked for miRNA binding to genes that encode DNA methylation machinery (none found).
other:
Zoe has also run Orthofinder to facilitate cross-species comparisons
Goals/To-Do (general):
Ensure I’m using the SMART framework (SMART: Specific, Measurable, Attainable, Relevant, and Time-bound)
Some goals are formulated to directly address one of the hypotheses listed above, and (for now) should be performed on A.pulchra data only.
“sRNA that bind to genes and are inversely correlated with expression - regulate those genes”
Use gene-miRNA interaction network to create some basic figures summarizing nature and extent of interactions (e.g. how many genes with how many miRNAs, how many genes does a single RNA usually interact with and vice versa, how many positive vs. negative relationships are there)“There will be a suite of genes that a sRNA regulates that have similar function”
Use functional annotation of the gene-miRNA interaction network to identify putative function of each miRNA based on function of associated genes.For predicted miRNA targets, compare functional composition of targets to functional composition of full reference.
(Basically, I want to see whether miRNAs preferentially regulate any functional classes).Formally name each miRNA, matching the names assigned during deep-dive when possible.
If an miRNA is formally described (e.g. mir-100), compare/contrast its described function with the putative function determined during (2)
“Some lncRNAs that act as sponges (bind) sRNA will impact gene expression sRNA regulate”
Use miRanda to predict miRNA-lncRNA binding and evaluate correlated expression of putative binding pairs to generate miRNA-lncRNA interaction network.“A subset of lncRNAs will regulate mRNA as evident by correlation”
Evaluate correlated expression of lncRNA and mRNA.“some lncRNAs may act as miRNA precursors – expect they will contain the full pre-miRNA”
Use BLAST to search for lncRNAs that fully contain pre-miRNA sequences.Using simulated WGBS data, prep pipeline to summarize methylation (whole genome methylation levels, GBM/distribution across genome features, etc.)
“Housekeeping genes: positive relationship between expression and GBM” and “plastic/environmentally responsive genes: negative relationship between expression and GBM”
Using simulated WGBS data, prep pipeline to delimit genes by function (housekeeping vs. responsive) evaluate relationship of both classes with gene body methylation.“siRNA modulate DNA methylation through RNA-directed DNA methylation by binding to genomic region and recruiting protein machinery to direct methylation”
Use miRanda/RNAhybrid to predict siRNA binding to the genome. Using simulated WGBS data, prep pipeline to test whether siRNA binding is significantly associated with methylationsiRNAs could also play a role in guiding DNA methylation to silence transposable elements
Use miRanda/RNAhybrid to predict siRNA binding to TEs. Evaluate correlated expression of siRNAs and TEs (expecting a negative relationship).
Multiomics?
At some point we’ll need to a) replicate these analyses on Peve and Ptuh, and b) run some cross species comparisons
Goals/Plan for Friday (1/31) meeting:
Clarify involvement/role of E5 molecular group members – ensure everyone who is/would like to be involved in deep-dive expression manuscript has something to be working on. Assign out goals/tasks if people
Define the overarching lens/focus of the manuscript
Once tasks are assigned to individuals and a rough timeline is selected, add tasks as Github issues
Add current progress to manuscript?