Genome Annotation Of Spruce
“So Zeno is most famous for his tortoise paradox. Let us imagine that you are in a race with a tortoise. The tortoise has a ten-yard head start. In the time it takes you to run that ten yards, the tortoise has moved one yard. And then in the time it takes you to make up that distance, the tortoise goes a bit farther, and so on forever. You are faster than the tortoise but you can never catch him; you can only decrease his lead.”
― John Green, The Fault in Our Stars
I have recently started the genome annotation of White Spruce - strain WS77111. For those who do not know: conifers are Gymnosperms and have diverged from the Angiosperm plants in an independent clade.
This is a picture that shows the plants evolution, on the final branches you can see the distinction between angiosperms and gymnosperms, both seed plants, which occurred 290 million years ago.
Langdale Lab - Department of plant sciences
The vast majority of the genome assembly and annotations that we have today is made by angiosperms species. At the current moment we only have draft reference genomes of several spices which is a great resource to start exploring the spruce genomics. There are also several genome annotations available on NCBI but those are bound to the current state of the assembly. If it’s fragmented and incomplete and also its annotation is partial. Being a pioneer is however something exciting!
We are a small Lab and our research topic is not primarly the genome annotation, as you may undertsand we are not a consortium for genome annoatation. The best method so far for our case is Maker which is a Perl wrapper combining other software. Being this my first time running a whole genome annotation I’ve realized that the most difficult part is to start the software and prepare all the opt/exe scripts. Maker has several dependencies and there are different perl libraries that need to be installed. Furthermore, in case of some type of errors it is quite complicated to figure out which software did not work. This was only the most apparent difficulty at the beginning of my test runs. There are many-many issues related to genome annnotation and my advise for anyone who wants to start such a big job is to check the intermediate output and never(-never) automate the pipeline without making a check on the results. Also: even if the software marks the genes as “high quality”, those still need to be validated for completeness, protein coding, length of introns etc.
The flavor of Maker that I run is Maker-P. It took me one week to understand what is that. There is only one page from Yandell lab that explains the pipeline, there is also a pubblication coupled to that but they’ve made some changes in the past years. It has more targeted software for repeat masking, detection for tRNA/snoRNA and Pseudogenes. Unfortunately, not all those tools are combined together in an unique scripts and the user needs to assemble the pipeline. Some of the features are already embedded in Maker2 but some others, such as the Pseudogenes pipeline, is totally independent and needs to be run after a first-step of Maker.
If you are also a student/staff or new to genome annotation and you need to annotate a highly divergent genome, I have some advises which I have learnt during my genome annotation project in the past months:
- Know your genome in terms of expected repeats (type, frequency, length etc.), gene structure and quality of assembly. This is fundamental to succeed
- Be patient and well-prepared. Maker is a pipeline that is supposed to run very easily but do not let Maker to do everything for you! Check, check, check. Most of the steps can be customized and adapted to the features of your genome
- Compile a list of high quality genes which you have validated with different methods or other tools. The top-notch can be used to train other models
- Compare your annotation to others performed by Maker to see how good is your job
- Put some love and passion in this work even if it looks like very technical. Remember that someone will use your annotated genes in future and that your work is IMPORTANT
Cheers,
References
Genome annotation - Maker