README.md

In arcolombo/junk: qusage: Quantitative Set Analysis for Gene Expression

SpeedSage Intro

qusage is published software that is slow for large runs, SpeedSage corrects for speed and efficiency at large orders

Bottlenecking of Functions

Qusage can improve the speed of its algorithm by minimizing the cost of computaiton.

changes calcIndividualExpressionsC

trading NA flexibility slows down qusage runs, but having the user input no NAs enforcing good input, this speeds up calcIndividualExpressionsC 2X

calculate Individual Expression Function

This test the local version which enforces no NA in Baseline or PostTreatment object, this reduces the flexibility. for non-paired end results:

min       lq     mean   median       uq      max neval cld
176.5508 224.7822 223.4475 226.2532 229.9082 242.5602   100   b
112.5428 116.8076 142.3476 163.1434 165.7608 170.5896   100  a

for paired end results:

 min       lq     mean   median       uq      max neval cld
 151.4904 175.8198 204.5262 191.4459 229.1922 332.8214   100   b
  129.6539 143.8261 166.9900 155.2507 191.1199 248.1528   100  a

for input expression matrix with NA enforced:

calcIndividualExpressionsC(testB,testPT)
Error in calcIndividualExpressionsC(testB, testPT) : 
  NA values are present in the expression matrix, please pluck out ...

makeComparison changes

The goal is to simplify costly functions by altering the design and control flow if possible. The second goal is to parallel compute the gene sets, rowSums command is one of the most computation time used, so if one can create a vector of eset Baseline, and PostTreatment and issue the rowSums in parallel this would reduce the compute time.

arcolombo/junk documentation built on May 10, 2019, 12:49 p.m.