Methods and systems for predicting cancer outcome
The invention provides a molecular marker set that can be used for prognosis of colorectal cancer in a colorectal cancer patient. The invention also provides methods and computer systems for evaluating prognosis of colorectal cancer in a colorectal cancer patient based on the molecular marker set. The invention also provides methods and computer systems for determining chemotherapy for a colorectal cancer patient and for enrolling patients in clinical trials.
1. A method for treating colorectal cancer in a human colorectal cancer patient, comprising:
obtaining a colorectal sample from the human patient;
measuring RNA transcripts corresponding to 7 genes having GenBank ID numbers N36176, AA149253, AA425320, AA775616, N72847, AA706226, and AA883496 in the sample; and
classifying said patient as having a poor prognosis by inputting the measurements to a prognosis predictor within a computer system comprising a processor coupled to memory, wherein the prognosis predictor comprises a neural network resident in the computer system, the neural network having been generated by
obtaining a training set of data from a population of colorectal cancer patients for whom marker profiles for the RNA transcripts corresponding to the 7 genes and prognosis outcomes are known,
determining, using the processor, one or more correlations between the marker profiles and a known prognosis of colorectal cancer in a human colorectal cancer patient, wherein determining a correlation comprises calculating a t-value,
weighting, using the processor, a correlation of the marker profiles based upon a value of the at least one marker profile within the population of colorectal cancer patients, wherein the weighting includes having the neural network start with a set of random weights, and
training the neural network with the training set of data to provide outputs indicative of a prognosis of colorectal cancer and adjust the weights to minimize error between the outputs to the known prognosis outcomes, wherein training the neural network comprises stochastic training, batch training, or online training;
providing, using the computer system, an output that includes the poor prognosis, wherein said poor prognosis predicts non-survival of a patient within said time period; and
treating the subject with a chemotherapy suitable for the poor prognosis.
2. The method of claim 1 , wherein said time period is 5 years.
3. The method of claim 1 , wherein measurement of each of said RNA transcript is obtained by a method comprising contacting a positionally-addressable microarray with nucleic acids from said cell sample or nucleic acids derived therefrom under hybridization conditions, and detecting the amount of hybridization that occurs, said microarray comprising one or more polynucleotide probes complementary to a hybridizable sequence of each said gene transcript.
4. The method of claim 3 , wherein said microarray is selected from the group consisting of cDNA microarray, ink-jet synthesized microarray, and oligonucleotide microarray.
5. The method of claim 1 , wherein said patient has an increased level of an RNA transcript of osteopontin or neuregulin 2 isoform 4 relative to the average level of said RNA transcript in patients without colorectal cancer.
6. The method of claim 1 , wherein said plurality of RNA transcripts comprises RNA transcripts of osteopontin and neuregulin 2 isoform 4, respectively.