Abstract: Methods for determining a tumor in a human is disclosed. Also disclosed are methods for identifying adenocarcinoma, and methods for identifying squamous cell carcinoma in a human tumor sample. In addition, methods for predicting prognosis of metastasis and survival in a human having a tumor is disclosed.

Abstract: The present invention provides methods using a gene expression profiling analysis (1) to determine whether a human sample is a tumor using a gene set containing nucleic acid sequences of SEQ ID NOS: 1-7, 8-17 or 1-17; (2) to identify whether a tumor tissue is an adenocarcinoma (using a gene set containing nucleic acid sequences of SEQ ID NOS: 15, and 18-21) or a squamous cell carcinoma (using a gene set containing nucleic acid sequences of SEQ ID NOS: 22-27); and (3) to predict the prognosis of survival and metastasis in humans with tumor (using a gene set containing nucleic acid sequences of SEQ ID NOS:19, and 28-42 or SEQ ID NOS: 19, 29, 31, 40, and 42), particularly for those humans who are at the early stage of lung cancer. The gene expression profiling is preferably performed by cDNA microarray-based techniques and/or Real-Time Reverse Transcription-Polymerase Chain Reaction (Real-Time RT-PCR), and analyzed by statistical means.