Skip to main content

Table 1 Prediction accuracy using k-NN on 92 white oak datasets of mixed sequence quantity based on \(d_{2}^{*}\) before and after bias adjustment for different query sizes, reference sizes, and different numbers of neighbors k used

From: Afann: bias adjustment for alignment-free sequence comparison based on sequencing data using neural network regression

Query sizeReference sizek = 1k = 2k = 3k = 4k =5k = 6k = 7k = 8k = 9k = 10
Before bias adjustment
1910.970.970.971.001.001.000.980.950.910.91
17750.980.980.960.990.960.980.960.950.910.91
32600.970.970.940.960.940.950.910.920.880.89
47450.950.950.930.940.910.910.880.890.870.88
62300.930.930.880.890.850.870.830.840.820.81
77150.840.840.770.780.750.740.690.700.670.65
After bias adjustment
1911.001.001.001.001.001.001.001.001.001.00
17751.001.001.001.001.001.000.991.001.001.00
32601.001.001.001.000.991.000.991.000.990.99
47451.001.000.991.000.990.990.980.990.970.98
62300.990.990.970.970.940.950.920.930.900.91
77150.960.960.920.930.870.870.810.790.740.70
  1. For each query sizes and reference sizes, the dataset was randomly split 100 times and an average prediction accuracy was calculated over 100 splits