May 26 - 30 notes

2014-05-24

ApTranscriptome

For working with DAVID database, needed gene IDs for high quality reference hits. Used the RefSeq protein database targeting Apis mellifera and Drosophila as preliminary examination by Sara showed these returned the most hits that were in the DAVID database.

The following script

downloads the RefSeq protein database,
makes a database alias for Apis mellifera and Drosophila as explained here but using taxid 7460 for Apis mellifera. note that the ‘gi.list’ file needs to be created and in the directory before running the script
runs blastx in parallel creating a file in BLAST archive format

#!/bin/bash

transcriptome="/home/jstantongeddes/research/ApTranscriptome/results/trinity-full/Trinity_cap3_uclust.fasta"

## download newest version of database using `update_blastdb.pl` script
update_blastdb.pl refseq_protein

## uncompress files
for filename in *.tar.gz
do
    tar zxf $filename
done

## make alias to A. mellifera and Drosophila
blastdb_aliastool -gilist ../amellifera_drosophila_gi.list -db refseq_protein -out refseq_protein_amellifera_drosophila -title "Apis mellifera and Drosophila Refseq Protein database"


# blastx using GNU parallel 
cat $transcriptome | parallel -j 8 --block 100k --recstart '>' --pipe blastx -evalue 0.01 -max_target_seqs 1 -outfmt 11 -db  refseq_protein_amellifera_drosophila -query - > ApTranscriptome_parallel_blastx_refseq_protein_amellifera_drosophila

I then created a table format file with 1 hit per query using blast_formatter:

blast_formatter -archive ApTranscriptome_parallel_blastx_refseq_protein_amellifera_drosophila -max_target_seqs 1 -outfmt 6 -out test6

Of the initial 105k transcripts, 29,080 had BLAST hits to the Apis mellifera & Drosophila RefSeq protein database. This is the file I then used for DAVID.

2014-05-27

ApTranscriptome

Figuring out how to use rDAVIDWebService…

Started Aster analysis user’s group!
Hacker’s guide to git
First meeting of bioinformatics bootcamp group
Update Silene aster analysis

2014-05-29

Centennial paper

Reading…

Jay et al. 2012 Molecular Ecology

applying SDMs to genetic clusters prone to two errors
statistical errors in assignment of individuals to clusters would not be accounted for
inaccurate modeling of admixture
ancestry distribution models account for these two problems
project ancestry coefficients on geographic maps
limit the number of clusters to < 4 …. why?

Sork et al. 2010 Molecular Ecology

range-wide characterization of genetic structure of California valley oak
patterns of population relatedness using PopGraph
significant connectivity among populations, especially in north. evidence for long-distance gene flow along foothills. central valley a barrier for direct east - west gene transfer in south
correlations between nuclear genetic variation and climate variation
strong association
region-specific bioclimatic models (SDMs)
size and proximity of predicted suitable habitat varies considerably among regions, from nearly complete overlap to separation of > 100 km

D’Amen et al. 2013 Global Ecology and Biogeography

examination of how species-level niche modeling can bias evaluation of conservation requirements in presence of intra-specific variation
10 mammal species in Africa with multi-gene phylogenetic trees
3 sets of ecological niche models (ENMs)
ENM-S: standard species-level ENM
ENM-L: for each lineage within a species
ENM-C: composite to summarize across each lineage
species-level models substantially over-predict range expansion

Espindola et al. 2012 Ecology Letters

investigate distribution dynamics of cold-adapted globeflower
characterized population structure using STRUCTURE
niche-based SDMs to predict future species and population distribution
compared hind-casted SDMs to phylogeographic
simulated

Pease et al. 2009

identify lineages of California mule deer using Structure
significant overlap of lineages at LGM
present-day genetic divergence appears to be recent, driven by ecological factors

2014-05-30

Meet with NG to discuss Centennial paper
Writing.
Working on DAVID analysis.
Meet with Scott Tighe at sequencing facility.
Went over protocol for bead purification and size selection using E-gels
very light DNA smear of post-PCR product. suggested we check PCR and do a titration with different # of cycles to confirm PCR is working