sigma_difference

The genome partitional data.(基因组分区数据)

The sliding window delta* profile data row. (CSV column names kept compatible with the legacy Site/Sigma/Similarity schema)

The word distribution r-scan result of one oligonucleotide word.

The sliding window delta* profile data row. (CSV column names kept compatible with the legacy Site/Sigma/Similarity schema)

One point of the strand composition asymmetry profile: (C - G) / (C + G) of one sliding window.

The codon signature profile of a gene collection: per-site nucleotide counts, site-specific dinucleotide counts, codon counts and the derived codon signature odds ratios {rhoXY(1,2)}, {rhoYZ(2,3)}, {rhoXZ(1,3)}, {rhoZW(3,4)}.

The codon usage profile of a gene collection (the relative codon frequencies and amino acid frequencies). This is the c(x, y, z) reference class data used by the B(F|C) bias measure (see CodonBiasMeasure) and the f^H(codon) reference set of the CAI (see CodonWeightTable).

CODON ADAPTATION INDEX (CAI, Sharp & Li) weight table.

This specifies a reference set of genes, H, chosen from among "highly expressed genes". Defining

    w(codon) = f^H(codon) / max_{synonyms} f^H(.)

as the ratio of the frequency of the codon to the maximal codon frequency in H for the same amino acid, the CAI of a gene of length L is taken as

    CAI(gene) = (w1 * w2 * ... * wL)^(1/L)

(the geometric mean, computed numerically through the log-average).

This is a two-step API: first build the w table from the reference gene set H (this class), then evaluate the target genes (see CodonAdaptationIndexEvaluation.CAI()).

(compatible rename of the legacy RelativeCodonBiases class, which mistakenly used each single ORF as its own reference set and thus always produced CAI values close to 1)

One gene prediction row of the alien gene 2D classification.

Load one sliding window delta* profile from its csv document.

Compile a set of pairwise delta* profile csv files (all created from the same query genome against different subjects) into one merged matrix csv document. The output file names cannot be changed by this requirement.

Compare one query genome against a directory of subject genomes: calculate the sliding window delta* profile of the query against every subject and export each profile as a csv file, then compile all of the profiles into one merged matrix csv document.

Sliding window delta* profile: the delta-difference between each local window of the genome and the comparison sequence.

PERFORMANCE: implemented with the incremental window count matrix (O(1) update per slide + O(16) distance per sample), and the comparison sequence signature is built only once.

Merge the exported partition-level delta* profile csv files with the two-way BLAST best hit rendering data.

Compile the codon usage table of every species gene collection. The reference set of each species is its own gene collection (for a real high-expression reference set use ToolsAPI.BuildCAIReference() with the ribosomal protein genes).

Compile the CAI w weight table of one reference gene collection and export the codon usage csv table. (legacy batch compilation entry)

Create the chromosome partitioning data from the two-way BLAST best hit data: for every partition tag, locate the corresponding ORFs on every subject genome and cut the partition nucleotide segments.

Save the partitioning data collection as a csv document.

Load the partitioning data collection from its csv document.

Load the chromosome partitioning entries from their csv document.

Calculate the pairwise delta* homogeneity matrix between all of the genome partition segments (the values are reported as delta* x 1000).

Create the partition data from a raw data frame: the start / stop columns define the segment loci on the given genome sequence.

Measure the homogeneity property of the genome partition segments against the dnaA-gyrB ruler segment of the reference genomes in batch.

Create a delta* distance matrix for a given sequence collection.

The delta* measure of difference between two sequences f and g (from different organisms or from different regions of the same genome): the average absolute dinucleotide relative abundance difference

 delta*(f, g) = (1/16) SUM |rho*XY(f) - rho*XY(g)|

where the sum extends over all 16 dinucleotides.

Using the DNA segment between the dnaA and gyrB genes as the reference ruler for the genome homogeneity measurement.

The rho* genome signature profile: the double-strand symmetrized dinucleotide relative abundance rho*XY = fXY / (fX fY) of all 16 dinucleotides, each reported with its six-level significance symbol (--- / -- / - / + / ++ / +++, thresholds 0.50 / 0.70 / 0.78 / 1.23 / 1.30 / 1.50).

The symmetrized tau* generalized odds ratio of one tetranucleotide:

 tau*(XYZW) = f(XYZW) f(YZ) / ( f(XYZ) f(YZW) )

A value far below 1 indicates a restriction site avoidance signal (e.g. the CTAG avoidance in M. jannaschii, tau* = 0.06).

The tau* tetranucleotide relative abundance profile of all 256 tetranucleotides.

List the rare and the frequent tetranucleotides of the tau* profile against the given thresholds (the restriction avoidance candidates).

Count the absolute occurrences of one oligonucleotide word in the genome sequence (e.g. the CTAG rarity report).

Locate all occurrences of one oligonucleotide word in the genome sequence (exact match, 0-based start positions).

Run the r-scan significance test on the spatial distribution of one oligonucleotide word (Dembo-Karlin 1992): detect the clustering, the overdispersion or the even spacing patterns against the random (Poisson) spacing expectation.

Build the site-specific codon signature profile of a gene collection: the odds ratios rhoXY(1,2) / rhoYZ(2,3) / rhoXZ(1,3) within the codons and rhoZW(3,4) at the codon junctions, distinguished from the global genome signature.

Build the codon usage frequency profile c(x, y, z) of a gene collection (the reference class C of the B(F|C) codon bias measurement).

The B(F|C) codon usage bias difference between two gene classes (the paper's formula [1]):

 B(F|C) = SUM_a p_a(F) SUM_{(x,y,z)->a} | f(x,y,z) - c(x,y,z) |

the per-codon absolute difference weighted by the amino acid frequency of the query class F. Note that this measurement is asymmetric: F is the query object and C is the reference ruler.

The B(g|C) codon usage bias difference of one single gene g against the reference class C.

Build the codon usage profile of the ribosomal protein gene set (the highly expressed reference set RP of the 2D threshold method).

Identify the alien (laterally transferred) genes with the 2D threshold method (the paper's B. subtilis thresholds by default):

B(g|all) > 0.42 and B(g|RP) > 0.45: alien gene - unlike both the average host gene and the host highly expressed genes;

B(g|all) > 0.42 but B(g|RP) < 0.45: highly expressed host gene.

The sliding window delta* profile of the genome against its own global average genome signature: the alien DNA that has not yet been ameliorated shows a signature deviating from the host average and therefore appears as peaks on the curve (e.g. the pathogenicity islands).

The (C - G) / (C + G) sliding window profile of the genome strand composition asymmetry: the leading strand favors purines (G > C), so the curve flips its sign at the replication origin oriC.

Locate the candidate replication origin (oriC) and terminus positions from the GC skew profile: the sign-flip sites of the cumulative skew curve. A single bidirectional replication origin produces one sign flip at oriC; genomes with many origins (e.g. some archaea) show no significant asymmetry at all.