论文部分内容阅读
基质辅助激光解吸电离质谱由于存在大量噪声,往往不能够较好地去除噪声。为了提高基线校正效果本文提出了1种新的基线校正方法——迭代自调整窗口极小值法。该方法将谱图分为多个窗口,按照吸收峰的分布自适应调整窗口宽度,将一部分窗口中的极小值用来构建漂移的基线。通过考察IMASW方法对模拟谱图的校正结果,发现该方法的性能明显好于固定窗口中值法和Tophat拟合方法,极小值策略很好地避免了有用吸收峰信息的损失,而且迭代使得IMASW方法对参数不敏感,可以用固定参数实现基线校正。将IMASW方法用于拟合100条MALDI-TOF血清多肽谱的基线,结果表明该方法快速有效,得到的基线光滑且不损失吸收峰信息,可以拟合各种变化形式的基线。考察拟合统计结果发现IMASW方法在校正基线过程中迭代3～8次,100条血清多肽谱平均需要运行4.57次得到最终结果,所耗费的CPU时间在0.15s～0.35s之间,平均使用0.3926s。显然IMASW方法非常快速,没有因为采用迭代策略导致效率和速度的降低。
Matrix-assisted laser desorption ionization mass spectrometry often does not remove noise well due to the large amount of noise present. In order to improve the baseline correction effect, this paper presents a new baseline correction method - iterative self-adjusting window minimum method. The method divides the spectrum into multiple windows, adjusts the width of the window adaptively according to the distribution of absorption peaks, and uses the minimum value of a part of the window to construct the baseline of drift. By investigating the calibration results of IMASW method on simulated spectrum, we find that the performance of this method is obviously better than the fixed window median method and Tophat fitting method. The minimum value method avoids the loss of useful absorption peak information well, The IMASW method is insensitive to parameters and can be baseline-corrected with fixed parameters. The IMASW method was used to fit the baseline of 100 MALDI-TOF serum polypeptide profiles and the results showed that the method was rapid and efficient, with a smooth baseline without loss of peak absorption information, and fitting to various variations of the baseline. According to the statistical results, we found that the IMASW method iterates 3 to 8 times during the baseline correction. The average serum peptide profile of 100 peptides needs 4.57 runs to get the final result. The CPU time spent is between 0.15s and 0.35s, with an average of 0.3926 s. Obviously the IMASW method is very fast and does not result in reduced efficiency and speed due to the use of iterative strategies.