/****** README **********/

five files included in archive named : posSet.fa, negSet.fa, random_walk.csv, dataset.csv, dataset_1.csv

The sequence data related to enhancer downloaded 
from ucsc table browser

posSet.fa	// sequence related to human developmental enhancer
negSet.fa	// sequence generated with similar gc content

/********************************************************
Both positive set and negative set file is combined.
-- first 1798 entries belong to posSet.fa
-- last 1798 entries belong to negSet.fa
*********************************************************/

random_walk.csv		// walk genrated using purine-pyrimidine model

Features are calculated using DNA random walk generated from purine-pyrimidine model

dataset.csv	// 11 features

Attributes :
1) min : minimum value in DNA random walk (drw)
2) max : maximum value in drw
3) median : median in drw
4) skew : skewness in drw
5) kurt : kurtosis
6) sd : standard deviation
7) var : variance
8) dfa : deterrent functional analysis
9) cor : correlation
10) hurst : hurst component
11) entropy : sample entropy

dataset_1.csv	// 39 features

Attributes :
1) min : minimum value
2) max : maximum value
3) median : median value
4) skew : skewness
5) kurt : kurtosis
6) sd : standard deviation
7) var : variance
8) dfa : deterrent functional analysis
9) cor : correlation
10) hurst : hurst component
11) entropy : sample entropy
12) asoc : absolute sum of changes
13) adf : augmented dickey_fuller
14) ac : auto correlation with lag 100
15) be : binned entropy with max bin 100
16) cam : count above mean
17) cbm : count below mean
18) flm : first location of maximum 
19) flmn : first location of minimum
20) llm : last location of maximum
21) llmn : last location of minimum
22) lsam : longest strike above mean
23) lsbm : longest strike below mean
24) mlfp : max langevin fixed point
25) mac : mean absolute change
26) ma : mean autocorrelation
27) mc : mean change
28) msdc : mean second derivative central
29) prdad :  percentage of recurring data points to all data points
30) prvav : percentage of recurring value to all values
31) rvntsl : ratio value number to timeseries length
32) srdp : sum of reoccuring data points
33) srv : sum of reoccuring values
34) vltsd :
35) ac_150 : autocorrelation with lag 150
36) ac_200 : autocorrelation with lag 200
37) ac_250 : autocorrelation with lag 250
38) ac_300 : autocorrelation with lag 300
39) lyap_r : lyapunov exponent

Sequence feature are calculated based on k-mer frequencies
featureDataset.csv  // combination of 2 to 6 k-mer in column and frequency of each sequence in rows