# Table 2 Given a referencerand a gene-disease pair <g,d>, CRFref estimates and integrates three measures: degrees ofconclusiveness,richness, andfocusofrwith respect to <g,d>

Factors Definition Type
(1) Length(r) $1 if length of r > AvgLe n a L ength of r AvgLen otherwise$ Degree of conclusiveness
(2) GeneTF(g,r) $1 if TF g , r > 5 b TF g , r 5 otherwise$
(3) DiseaseTF(d,r) $1 if TF d , r > 5 TF d , r 5 otherwise$
(4) Gene@Title(g,r) $1 if g appears in the title of r 0 otherwise$
(5) Disease@Title(d,r) $1 if d appears in the title of r 0 otherwise$
(6) Gene@Ending(g,r) $LastPos g , r c length of r$
(7) Disease@Ending(d,r) $LastPos d , r length of r$
(8) NotGeneNum(g,r) $1 if G ' > 5 G ' 5 otherwise G ' = g ' | g ' ∈ G − g and appears in r d$ Degree of richness
(9) NotDiseaseNum(d,r) $1 if | D ' | > 5 D ' 5 otherwise D ' = d ' | d ' ∈ D − d and appears in r e$
(10) NotGene@Title(g,r) $1 if there is a gene g ' ∈ G − g that appears in the title of r 0 otherwise$ Degree of focus
(11) NotDisease@Title(d,r) $1 if there is a disease d ' ∈ D − d that appears in the title of r 0 otherwise$
(12) NotGene@Ending(g,r) $Ma x g ' ∈ G − g and appears in r LastPos g ' , r length of r$
(13) NotDisease@Ending(d,r) $Ma x d ' ∈ D − d and appears in r LastPos d ' , r length of r$
1. [a]AvgLen is the average length of references.
2. [b]TF(x,r): Term frequency of x in r.
3. [c]LastPos(x,r): The last position of x in r.
4. [d]G: Set of gene names in HUGO Gene Nomenclature Committee (HGNC).
5. [e]D: Set of terms in MeSH class of C04 to C26, with ‘disease’ and ‘syndrome’ removed.