# Filtered and denoised reads

**URL:** https://forum.qiime2.org/t/filtered-and-denoised-reads/9471
**Category:** User Support
**Created:** [May 1, 2019, 1:22pm UTC](https://forum.qiime2.org/t/filtered-and-denoised-reads/9471 "2019-05-01T13:22:24Z")
**Posts on this page:** 1
**Showing post:** 4

<div class="post-metadata">

### Author: ![yakshi.ortiz](https://forum.qiime2.org/user_avatar/forum.qiime2.org/yakshi.ortiz/32/4170_2.png) [@yakshi.ortiz](https://forum.qiime2.org/u/yakshi.ortiz)
#### Post date: [May 2, 2019, 8:30pm UTC](https://forum.qiime2.org/t/filtered-and-denoised-reads/9471/4 "2019-05-02T20:30:23Z")

</div>

Hi Yogesh,

If I understand right, what the **denoising step** does is to correct sequencing errors at nucleotide level. This is why you will see very low to no loss at this step. The **filtering step** removes sequences based on the selected _max-ee_ value. The max-ee is the sum of probabilities that each of your nucleotides were called incorrectly (this is based on the phred score of each of your nucleotides in a given sequences). If that sum is more than the selected max-ee, then your sequence is discarded. You can read more at this [link](https://www.drive5.com/usearch/manual/exp_errs.html).

I hope that helps you a bit!  
Yakshi

---

_[View the full topic](https://forum.qiime2.org/t/filtered-and-denoised-reads/9471)._