The paper presents a method that integrates ontology-based semantic tagging, information extraction and data mining together, automatically learning protein interaction patterns based on a few user seed tuples, and then extracting new tuples from biomedical web documents based, on the discovered patterns. A system, called SPIE (Scalable and Portable Information Extraction), is implemented and tested on the PuBMed to find the chromatin protein–protein interaction and the experimental results indicate the approach is promising.

Shows FACT destabilizes nucleosomes so that H2A-H2B dimers are removed. This was shown in vitro with recomb hFACT. -they show SSRP1 (a fusion of Pob3-Nhp6) interacts with H3/4 but not H2A-H2B. -showed hFACT does have chaperone activity (deposites core histones onto DNA in vitro.

-paper solved structure of N term domain of Spt16 in FACT. This domain of spt16 and Pob3-M have overlapping rokes in binding the C-term extension of H2A. Therefore FACT has additive contacts with nucleosomes. -H2A binding site is in POb3-Q308 region previously described as binding RPA. -trypsin treated nucleosomes that have lost their tails no longer bind yFACT in vitro. -Spt16-Pob3 binds H3 and H4 tails in vitro by SPR. The Spt16NTD is not required for this interaction. The complex does not bind H2A tail. Didn't test H2A globular domain. This paper never tests H2A C-term extension mutants for FACT binding. Only tested genetically by synthetic interactions between FACT mutants and H2A mutants. Not very convincing that a physical interaction exists b/w H2A C term and FACT.

-this paper solves the crystal structure for Pob3, a member of FACT. -pob3 has an unusual double PH domain. A mutant w/in this domain that causes defects in replication is rescued by a mutation in RPA a ssDNA binding factor. This indicates that FACT coordinated with RPA during DNA replication. -The binding surface is on the N terminal PH domain. This domain is much less conserved between Pob3 and Rtt106. The C term is v conserved (likely the site of histone binding). -FACT and RPA display a weak physical interaction. -FACT alters nulceosomes independent of ATP therefore it's referred to as a reorganizer rather than a remodeler. -human FACT was purified by its ability to promote RNA pol II transcription on chromatin templates in vitro. -there are 6 PH domain subgroups. Pob3 PH domains are distinct and therefore make up a 7th group. -PH domain suggests it binds to a ligand but it doesn't suggest what the ligand it (lipid, peptid, etc). -PH domain: 7 stranded anti-parallel beta barrel that is capped at one end by a helix. -to identify pob3 regions involved in DNA replication they performed random mutagenesis on all of pob3, and screened for mutants that were sensitive to HU and were able to grow at 37 C. This indicates replication defects and that weren't a result of protein instability. -screen revealed Q308R and Q308K. This is in an external loop of the N term PH domain. -found supressor mutation in RPA. H2B: previously reported that replication defects caused by pob3 alleles can be rescued by increasing the ratio of H2A/H2B to H3/H4. This rescue is thought to be associated with transcription rather than replication. -pob3Q308K strains displayed growth defects when HTA2-HTB2 was o/e compared to HTA1/HTB1. This growth defect was exacerbated 75mM HU (a condition permisable to pob3-Q308K. -this trend was also observed in the RPA mutant background. Overexpressing H2B copy to was more detrimental that o/e copy 1. This may be because copy 1 promoter self regulates and therefore not as much protein was expressed. -They think maybe because copy 2 isn't transcriptionally repressed by the Hir proteins. The mutant pob3 has diminished tolerence for increased expression of H2A/H2B to H3/H4.

Deletion of pob3, a component of pombe FACT, is viable but has similar 'lagging chromosomes' phenotype to RNAi and heterochromatin pathway mutants. They also show that siRNA pathway is fine, but mutants have reduced Swi6 association with centromere.