PM05:ARTIFICIAL  NEURAL NETWORK CLASSIFIER (ANNC) BASED ON PREDICTIVE

            MUTATION GENERATOR (ANNC-PMG)

ANNC  has as its goal to improve diagnosis of  NSCLC based on  sample patients’ data with microdeletion mutations (exon 18,19,20) and nucleotide conversion(exon 21)  extracted from online EGFR mutation database,  and  samples  data with prediction microdelition mutations generated in our own generator.  We have developed an integrated software suit  based on module for preprocessing data (extraction, encoding, and  normalization), module for mutations generator based on satisical exon microdelition, module for training/learning of ANN, and module for postprocessing (classification, and  evaluation). Experiments have  been done on eleven different training/learnig algorithms in  combination with different number of cells, layers, and activation functions. The best results have been achieved  with cascade-forward backpropagation algorithm based on Levenberg-Marquardt learning mechanism, including best performance (error 5e-031) with the minimum epochs ( training iterations 6) and the  regression fit curves (trainig,validation and testing  R=1). The whole set have been divided in 700 training pairs  and 411 pairs which serve for validation. Through free selection of validating pairs the classificator has successfully divided the positive cases (affected by illness) and the negative (healthy)  ones. But this approach  to gene based cancer classifications uses the data about different exon mutations from public databases, where the exact classification is possible only  if in the case of a  new patient the exact match is found in  database.

     Figure 1. Block structure of ANNC in the process of training

It is not enough to perform the classification of patients into one already well-known statistical group, but if there is a patient with a new mutated exons, our algorithm need to determine the position and number of mutated nucleotides. Because of that we want to develop more powerfull mutation  prediction  generator  for microdeletion  mutations that take place over the  nucleotides in distributed stohastic order related to position and numbers of mutations.

EXON 18

2123 

1.063 x  1037

EXON 19

299

6.33  x    1029

EXON 20

2186

9.80 x    1055

EXON 21

274 ∙ 335

9.45  x   1038

To develop prediction mutation generator that cover all  combinations of distributed nucleotides mutations related to positions in exons and number of mutated nucleotides would be very long process, and is not in line with human life. Because of that we  look for a solution in new technologies (parallel multiprocessing cores), and the application of new models based on fusion of artificial intelligence methods.