Classification of protein profiles using fuzzy clustering techniques: an application in early diagnosis of oral, cervical and ovarian cancer

 Present  study  has  brought  out  a  comparison  of PCA  and  fuzzy  clustering  techniques  in  classifying  protein
profiles  (chromatogram)  of  homogenates  of  different  tissue origins:  Ovarian,  Cervix,  Oral  cancers,  which  were  acquired using HPLC–LIF (High Performance Liquid Chromatography- Laser   Induced   Fluorescence)   method   developed   in   our laboratory. Study includes 11 chromatogram spectra each from oral,  cervical,  ovarian  cancers  as  well  as  healthy  volunteers. Generally  multivariate  analysis  like  PCA  demands  clear  data that   is   devoid   of   day-to-day   variation,   artifacts   due   to experimental   strategies,   inherent   uncertainty   in   pumping procedure which are very common activities during HPLC-LIF experiment.  Under  these  circumstances  we  demonstrate  how
fuzzy clustering algorithm like Gath Geva followed by sammon mapping   outperform   PCA   mapping   in   classifying   various cancers from healthy spectra with classification rate up to 95 % from  60%.  Methods  are  validated  using  various  clustering indexes   and   shows   promising   improvement   in   developing optical pathology like HPLC-LIF for early detection of various cancers  in  all  uncertain  conditions  with  high  sensitivity  and specificity.