9:40 A – 7:15 P, 8:00 P – 8:30 P,
remotely, 9:00 P – 10:30 P
Coding
- see additional project notes for today.
- Got OligoArray working on Cajal, figuring out ideal batch size and memory limits:
- Processes eat ~1.8 GB memory for 1 GB memory allowed (used by BLAST)
- Running 15 processes easily shoots through the memory ceiling and causes errors
- Running 6 processes uses at most 1/4th of Cajal CPU and 10 GB of memory (allowance is 1 Gb/process).
- 1 GB of memory causes a number of BLAST processes to crash. This may slow analysis, lead to lost data?
- Relaunched again at 11:15 A with 6 processes and a 3Gb memory limit.
- Note, each core saves to a different folder (indexed from 0) inside the target directory.
- Analysis currently (12:30A) stalling at
Start hg19_knownGene_uc001agc_e2_e1_pt2
. Only 4% of CPU active, 15 GB ram in use. - Maybe less memory would be better — make BLAST crash sooner when it reaches a ridiculous number of hits. This seems to be the major hang-up.
- would be good to right a pick-up where-left-off script.
- Running finally on Odyssey
- Not supposed to submit large batch jobs directly from odyssey node. Supposed to use batch launch option to appropriate queue. Not sure how to do this.
OligoPaints
Analysis of efficiency
- See qubit notes and concentration table
Set up Nicking Reactions
- samples 1-3
- Combine DNA from glycogen and no glycogen precipitation = 90 uL
- Add 11 uL buffer 2, 10 uL RE, incubate at 65C 5 hrs. heat denature 80C 25 min, then 4C
- samples reacting at 2:00 P.
- samples moved to 4C at 7:00 P
Set up T7 Exonuclease Reactions
- samples 4 and 5
- 15 uL DNA
- 5 uL buffer 4
- 5 uL T7 exonuclease
- 25 uL ddH2O
- take 1 uL aliquots as T=0. store at 4C
- samples reacting at 37C, 2:15 P.
- samples EtOH precipitating with glycogen added O/N at -20C
- 1 uL of each sample saved for diagnostic purposes, diluted 1:10.
Write Outline for Chicago Talk
- started making slides
Update online Banff Confirmed participants
- didn’t manage to do this today.