1/15
Looks like no tags are added yet.
parts list
the cells it has and the genes those cells express
Then you can figure what these cells/genes do by perturbing the system
Getting that information is not trivial, especially if no molecular tools exist for your species
Model Organisms
Over 2/3’s of the dev bio literature comes from just 7 organisms
Because they’re experimentally tractable (easy to test things in)
• Ease of growth, observation, or manipulation
• Tools, infrastructure, and resources built over decades

Most animals are not like this, but in the last ~20 years, fallingvsequencing costs and new tools have made it feasible to work in many more species
• Accessibility to genome sequencing
• Single-cell sequencing
• Improvements in microscopy
• CRISPR-Cas9
The biggest bottleneck to molecular investigation
lack of a well annotated reference genome
Without it, you’ll have to work really hard to label a gene and find out where it’s expressed
Genome Sequencing
The cost of sequencing a whole genome dropped significantly in the late 2000s
Now we can make very workable reference genomes for novel species for ~$7-10K, and solid ones for ~$10-25K
Single-Cell RNA Sequencing
scRNA-seq provides a way to identify and characterize cell types in a sample based on gene expression profile without needing to know gene markers ahead of time
1. Cells are dissociated and attached to barcoded beads in a microfluidic system
2. RNA in each bead/cell droplet is tagged so we know where each molecule came from
3. Sequence the nucleotides and get a cell x gene table
4. Group cells into clusters based on similarity
cell atlas
You can see what genes are in which clusters and use knowledge about those genes to name the clusters

Lmitations to Single-Cell RNA Sequencing
• Clusters are hypotheses. Cluster resolution is a parameter you control
• Clusters can be technical artifacts: empty droplets, droplets with multiple cells, over-splitting data
• Annotating clusters depends on existing knowledge, which might not exist for your genes
• Lose spatial information
Pseudotime
orders cells based on similarity (not real time), to infer differentiation trajectories

Spatial Transcriptomics
A variant of the technology, spatial transcriptomics, is able to capture where genes are expressed in a tissue, overcoming that loss in traditional scRNA-seq
Fluorescent Microscopy
Light-sheet microscopy: gentle, fast live imaging of whole embryos, was used to track 92% of cells through zebrafish gastrulation
HCR/FISH
Fluorescent in situ hybridization (FISH) detects localization of RNA using fluorescent probes
HCR improves upon FISH by increasing signal to noise ratio and allowing for easier multi-gene stains
Found it
So, with access to a sequencing core, a really nice microscope, and careful experimental design you can have a full cellular and genetic map for development of your non-model organism
CRISPR
CRISPR-Cas9 is a system identified as a viral defense mechanism in bacteria
An endonuclease (Cas9) is loaded with a guide RNA that directs it to DNA that shares that sequence like a heat seeking missile
Then it cuts that DNA like fancy molecular scissors
All you need is the DNA sequence and you can design a guide RNA for it
When you cut DNA with crispr…
Lose it
If there’s no repair template, the cell will try to fix the break but introduce random insertions/deletions, which will usually make a gene non-functional (knockout)
Move it
If there is a repair template, the cell will (rarely) fix the break with that template and incorporate any new DNA that comes with it (knock in)
There are also plenty of other ways Cas9 has been modified to do lots of other things, and still, all you need is the DNA sequence to direct it there
CRISPR Screens
If you don’t know what gene to target, no problem! Target a bunch all at the same time! This is a CRISPR screen.
A modified version of a forward genetic screen (break everything and see what gives a phenotype)