I still don't really understand what you're asking. 行く is いく is iku is 行く. I don't know what you're getting at talking about deriving romaji from symbols and things. い is is equivallent to 'i', く is equivallent to 'ku'. 行 is a kanji and it's pronunciation varies depending on what word it's used in, but in this one, it's covering the い part of the word いく, thus 行く.
The only thing I can guess is that you don't know that the hiragana and katakana are phonetic systems, and may misunderstand a couple things about the kanji as well. If this is the case, go to
http://en.wikipedia.org/wiki/Japanese_writing_system and give a brisk read-through of the main page and the three linked pages for kanji, hiragana, and katakana.
Also, I can't tell, you may already understand this, but it may be worth pointing out that, no, people don't memorize that 行く is iku, but rather that 行く is いく. Romaji is merely an outside tool used for a few misc purposes, one of which is early (foreign) Japanese language instruction. The basic phonetics of the language are in kana, which is one of the main reasons comments on deriving the romaji from kana confuse me. I could possibly understand what you meant if you were asking about how kanji hold different pronunciations in various contexts, but kana are simply what they are, and there are no tricks to them. 行く would always be (something)ku because く is ku and that's that.
Oh standard disclaimer: Don't be put off by my tone or whatever. I often come off differently than I intend to while explaining things. I am legitimately hoping that you find the answer to your question here.