Drosophila Annotation Goals: Final Presentation and Written Reports

The abstract should provide a brief statement of the goals and a report of your findings. This should be similar to and about the same length as a typical abstract found at the beginning of a scientific paper (~300 words). Your abstract should be a summary of your results, NOT the process by which you achieved those results. For the D. bipectinata annotation project, be sure to include the size of the project, number of Genscan predictions, protein-coding genes documented, and other verified features (e.g., novel repeats, non-coding RNAs) or unusual findings (e.g., changes in exon structure, changes in the number of isoforms), etc. Be specific in reporting what you have found and your resulting conclusions.

One to three paragraphs: Why are we studying this problem? Why is comparative genomics a powerful approach for analyzing the features of genes and chromosomes? What kinds of results are we generating right now, and what do we hope to derive in the future aggregate analysis? Try to keep this section concise and informative, but give some context — what are the goals of the overall study, and what are your goals in particular. Include a figure covering the entire sequence of your project(s) that indicates the size and position of each putative feature you will investigate. A screenshot of the GEP UCSC Genome Browser for your project region may provide a good starting point for this figure. Box and number the features that you will investigate/report on for ease of communication.

This section will be significantly larger (reflecting the number of features in your project), and it should be subdivided with a sub-section for each putative feature. You should present here a more detailed analysis of any genes, pseudogenes or partial genes you find in your project. Use the GeMoMa predictions (or similar) as an organizer for your presentation. Remember that while in D. melanogaster pseudogenes are rare, this may not be the case in these regions of high expansion so keep an eye out for evidence suggesting the presence of pseudogenes in your project. Because the projects are partitioned into sub-regions, you may find only part of a gene at one end or the other of your project. Also be aware that because we are annotating the draft whole genome assemblies, sequencing errors may also occur. Note that ab initio gene finders may miss some candidate genes, so do not limit your investigation to these predictions; look at the other evidence tracks (homology search with D. melanogaster, RNA-seq, etc.) and examine the orthologous region of the D. melanogaster genome as annotated by the curators at FlyBase. In particular, Genscan (and other ab initio gene predictors) will not predict non-coding RNAs (ncRNAs); be sure to investigate those posted for D. melanogaster, looking for conserved sequences in your species (e.g., via a blastn search of the non-coding RNA against your contig). When investigating a feature, you should use all of the information available to you: gene predictions, results from BLAST searches, RNA-seq data (e.g., splice junction predictions and assembled transcripts), and conservation tracks for the other Drosophila species (see the walk-throughs and practice examples).

For each gene in your D. bipectinata project, start by confirming the D. melanogaster orthologue; include the FlyBase ID and the name of any D. melanogaster gene that you consider likely to be orthologous in your report. After making approximate assignments for each exon using the "Align two or more sequences" functionality in BLAST, determine the exact location of each coding exon using all available evidence. Construct a model for each putative isoform. Use the Gene Model Checker to confirm your basic decisions; be sure to check both the dot plot and the peptide sequence alignment comparison to D. melanogaster. Every difference between your model and the D. melanogaster orthologue represents an assumption of evolutionary change that underlies your gene model. Based on the GEP annotation protocol, the best gene model should minimize these assumptions; as such, you should be able to account for every difference and convince yourself that there is no other acceptable gene model with fewer differences. [If possible, these differences should be supported by biological evidence (e.g., RNA-Seq data) or sequence conservation with other Drosophila species.] The dot plot and peptide sequence alignment, with a discussion of important differences (for each isoform), should be part of your report.

Note that information from high-throughput sequencing of RNA, the RNA-Seq track and its derivatives, will be very useful. RNA-Seq sequences were generated by producing many short reads (100-125 bases) from expressed RNAs (in many cases using total RNA from embryos or from adult females or males) using the Illumina sequencing technology. These reads have been mapped back to the genome sequences and the number of reads that overlap each base plotted in the "RNA-Seq Coverage" tracks. In addition to using the RNA-Seq results to help you with your annotations, you should also compare the RNA-Seq data with your final gene model. Relevant questions include: How well does the RNA-Seq coverage track correspond to the locations of the individual exons in your annotation? Can you reliably use RNA-Seq coverage to identify untranslated regions (5' and 3' UTRs) for each of your genes? Of all of the introns in your gene model, how many were supported and how many were unsupported by results in the splice junction and transcript tracks? Be sure to investigate all regions that have high RNA-Seq coverage, especially in regions that do not correspond to the genes you have annotated. (However, some of the regions with high RNA-Seq coverage could be spurious if they overlap with transposon remnants or other repeats.) Note that genes that are expressed only at low levels or in a restricted set of tissues or developmental stages may not display significant RNA-Seq data for our species. You can assess the expression status of the orthologous gene in detail from the data available for D. melanogaster on https://flybase.org.

As described above, you should provide a detailed description of your general approach with the description of the first gene discussed in the paper. Be sure to include screenshots documenting the following:

Include the final model in a table, showing the exact exon coordinates, size, percent identity with D. melanogaster, reading frame, phase of the donor and acceptor sites. While you should certainly include a dot plot and protein alignment from Gene Model Checker (GMC), a screenshot of the GMC checklist (i.e., green check marks) should not be shown if every test passes. Figures like this do not hold informational content that cannot simply be described in the text (i.e., "all tests passed in GMC") and is evident from the dot plot and protein alignments. However, you may include the GMC checklist as an image if it contains any warnings of errors (e.g., non-canonical splice donor). In all cases any warnings or errors must be discussed in the main text and presented along with the evidence you used to support these exceptions. Be sure to include a detailed justification in your report if the proposed gene model for your species differs substantially from the D. melanogaster orthologue. There will be differences — evolution happens! But you should be able to explain any substantial gaps in the dot plot, particularly large vertical or horizontal gaps (which correspond to large insertions or deletions compared to the D. melanogaster orthologue, respectively).

For subsequent genes, show verification of the orthologue, the table of exons described above, and the Gene Model Checker output (dot plot, and protein alignment). Visual evidence of your decisions on start, stop, and splice sites will be needed only if the decision was difficult because the evidence is ambiguous, the circumstances are unusual, or if it results in substantial changes compared to the D. melanogaster orthologue.

You will need to report the exact position of each coding exon in every gene as a table in an Appendix, and as a separate file in the format described below. Remember when trying to precisely place intron/exon boundaries that introns (almost) always start with the two bases, 'GT', and end with 'AG'. Non-canonical splice sites (e.g., GC splice donor site) should only be used if they are supported by conservation with D. melanogaster, or conservation with neighboring species (e.g., see the "Drosophila Conservation (28 Species)" track in the genome browser for D. melanogaster), or with RNA-Seq data (e.g., splice junctions).

As we have discussed, we would like you to search for the TSS position(s), narrow TSS search region, and the wide TSS search region for the genes in your D. bipectinata region using various D. ananassae derived query sequences. Relevant tracks in the D. ananassae genome to find relevant query sequences will include RNA-seq data, the strand-specific RAMPAGE data and the ATAC-Seq peaks identified by MACS2. Start by an examination of the RNA-seq and RAMPAGE data to define the D. ananassae narrow TSS region and if possible the TSS position. Follow this by searching for viable regions of significant ATAC-Seq signals to identify a TSS wide region. For each of these regions, extract the relevant base sequence from D. ananassae and use that in a blastn search of your project region using the more sensitive parameters. Report all results, positive or negative, irrespective of outcome. Be sure to interpret any results you get! If you do get blastn alignments do they make sense? Are they on the correct strand and in a plausible position to actually be the TSS in D. bipectinata? (Success with this strategy has proven to be better than we initially hoped!)

You should complete the TSS analysis on at least two genes. As with the annotation of the protein genes, give a detailed analysis in the first instance with sufficient detail to allow the reader to reproduce the analysis. For the second and any subsequent genes, report a succinct summary of the analysis steps and discuss your interpretation and conclusions, using a table or figure (better) to present your results.

The analysis of repeats is most interesting with respect to the questions of heterochromatin and euchromatin and the role of repeats in chromosomal expansion. Include in your report the Repeat summary statistics, repeat class distributions, most common transposons (by overall coverage in your region), and the more detailed analysis of the one RepBase transposon you claimed from the "Repeat Claim List". (The "Repeat Analysis Worksheet" can help you keep track of these statistics.) Comment on the results: do you see many different repeat families, only a few? Give exact numbers on the distribution of repeats found in your region and comment (either here or in the final discussion section) on what your results say about the mechanism of chromosomal expansion in D. bipectinata.

Generate a simplified map of your D. bipectinata project region, including genes and large (≥ 500bp) repetitious elements. Analyze a region of 5 genes from your project. Using one of the D. bipectinata F element genes in your project region, identify the location of the corresponding ortholog on the D. melanogaster F element. Generate a map of the D. melanogaster genome centered around this gene which covers approximately the same number of protein-coding genes as your D. bipectinata project region. Note which Muller element each gene is on in D. melanogaster.

Using these two maps, compare the relative gene order and orientation in D. bipectinata and D. melanogaster. Remember that the orientation of your D. bipectinata project is arbitrary (based on computer generated files), so you should focus on the relative orientations of the genes when analyzing synteny. (The project region is considered to be syntenic with D. melanogaster if the reverse complement of the D. bipectinata project has the same gene order and orientation as D. melanogaster.)

If synteny has been preserved, the genes in your project will all be from the same region of the D. melanogaster genome, and your comparison will be of one map to the other (two horizontal lines). If the regions are completely syntenic, the genes will be in the same relative order and orientation in the two species. If synteny has not been preserved, please include a comparison based on each gene in your project to the same gene and flanking regions in D. melanogaster (several horizontal lines, stacked up). Look for evidence of events (such as inversions, transpositions, etc.) that could have easily occurred during evolution that could explain the changes in synteny. If there is no set of simple events, then consider the changes as "complex" and report as such.

Since the D. bipectinata F element is substantially larger than the D. melanogaster F element, we would like you to check your region for expansion. Consider your project region as a series of intergenic, 5' UTR, CDS, and intronic subregions (you may ignore 3' UTR’s). For the purposes of this analysis, use the one isoform which has the largest number of CDS’s for each gene. Calculate the mean size of each type of subregion and compare these to the mean size of the orthologous regions from D. melanogaster. Finally interpret your results. Which of these subregions have expanded? Some of them? All of them? How do the expansion rates compare? See the "Beyond annotation lecture" for details and advice.

As time permits, we encourage you to explore one of the genes or features in your project in more detail. What is the presumed function(s) of one of the genes? Can you generate evidence in support of these functions? Are key regions of the predicted protein conserved? Looking at the "Gene Ontology" and "Protein Domains" sections of the FlyBase report, and the Clustal Omega analysis results may be informative. Is this gene part of a known biological pathway in D. melanogaster (see the "Pathways" section of the gene report)?

One to three paragraphs. Start with a "final map" in this section. Use custom tracks to show the entire project region with all the final annotations of all CDS-based gene models for all members of the group. Point out the genes and TSS’s you were directly involved in annotating. Summarize the accomplishments of the group. In subsequent paragraphs, put your work in context. How do your findings support or differ from prior analyses? Refer in particular to the results of Leung et al., 2017 and questions raised in the current grant proposal.